** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

Dash-html-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

Dash-html-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

Dash-html-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

Dash-core-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

Dash-html-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

Dash-core-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

** DISPUTED ** Py throughout 1.11.0 allows remote attackers to conduct a ReDoS (Regular expression Denial of Service) attack via a Subversion repository with crafted info data because the InfoSvnCommand argument is mishandled.
https://github.com/pytest-dev/py/issues/287

Py 1.10.0 includes a fix for CVE-2020-29651: A denial of service via regular expression in the py.path.svnwc component of py (aka python-py) through 1.9.0 could be used by attackers to cause a compute-time denial of service attack by supplying malicious input to the blame functionality.

Affected versions of Idna are vulnerable to Denial Of Service via the idna.encode(), where a specially crafted argument could lead to significant resource consumption. In version 3.7, this function has been updated to reject such inputs efficiently, minimizing resource use. A practical workaround involves enforcing a maximum domain name length of 253 characters before encoding, as the vulnerability is triggered by unusually large inputs that normal operations wouldn't encounter.

Affected versions of the idna package are vulnerable to Denial of Service due to an incomplete fix for CVE-2024-3651 that still allows specially crafted inputs to consume significant resources during encoding. The idna.encode() function invokes the valid_contexto validator on every label before applying length-based rejection, so payloads such as long repetitions of the Arabic-Indic digit U+0660 or sequences of the Katakana middle dot U+30FB followed by a CJK character cause valid_contexto to perform extensive context-rule processing across each character. A remote attacker who can supply domain-name input to an application that calls idna.encode() without first enforcing the 253-character DNS length limit can submit arbitrarily large strings that drive the validator to exhaust CPU time, resulting in Denial of Service through resource consumption.

Earlier versions of Dash and its components are susceptible to an XSS vulnerability, specifically through the manipulation of the href attribute in a tags by an attacker. This flaw could potentially allow an authenticated attacker to access or manipulate user data and tokens, assuming the ability to store and present manipulated views to other users. The vulnerability notably requires the presence of user input storage mechanisms within Dash applications to be exploitable. Further details are covered under CVE-2024-21485.
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user. See CVE-2024-21485.

Dash 1.21.0 updates its dependency 'Plotly.js' to v2.2.1 to include a security fix.

Dash 2.15.0 validates the URL to prevent XSS attacks identified on the 'dash-core-components'. 
https://github.com/plotly/dash/pull/2732

Dash 1.20.0 fixes a potential XSS vulnerability by starting to validate callback request fields.
https://github.com/plotly/dash/pull/1546

Affected versions of the Flask package are vulnerable to Information Disclosure due to missing cache-variation headers when the session object is accessed via certain code paths. In Flask’s session handling, accessing flask.session is intended to set a Vary: Cookie response header, but session key-only access patterns (such as using the Python in operator to test for a key without reading or mutating session values) can bypass the logic that adds the header.

Flask 2.2.5 and 2.3.2 include a fix for CVE-2023-30861: When all of the following conditions are met, a response containing data intended for one client may be cached and subsequently sent by the proxy to other clients. If the proxy also caches 'Set-Cookie' headers, it may send one client's 'session' cookie to other clients. The severity depends on the application's use of the session and the proxy's behavior regarding cookies. The risk depends on all these conditions being met:
1. The application must be hosted behind a caching proxy that does not strip cookies or ignore responses with cookies.
2. The application sets 'session.permanent = True'
3. The application does not access or modify the session at any point during a request.
4. 'SESSION_REFRESH_EACH_REQUEST' enabled (the default).
5. The application does not set a 'Cache-Control' header to indicate that a page is private or should not be cached.
This happens because vulnerable versions of Flask only set the 'Vary: Cookie' header when the session is accessed or modified, not when it is refreshed (re-sent to update the expiration) without being accessed or modified.
https://github.com/pallets/flask/security/advisories/GHSA-m2qf-hxjv-5gpq

Numpy 1.22.0 includes a fix for CVE-2021-34141: An incomplete string comparison in the numpy.core component in NumPy before 1.22.0 allows attackers to trigger slightly incorrect copying by constructing specific string objects. 
NOTE: the vendor states that this reported code behavior is "completely harmless."
https://github.com/numpy/numpy/issues/18993

Numpy 1.21.0rc1 includes a fix for CVE-2021-33430: A Buffer Overflow vulnerability in the PyArray_NewFromDescr_int function of ctors.c when specifying arrays of large dimensions (over 32) from Python code, which could let a malicious user cause a Denial of Service. 
NOTE: The vendor does not agree this is a vulnerability. In (very limited) circumstances a user may be able provoke the buffer overflow, the user is most likely already privileged to at least provoke denial of service by exhausting memory. Triggering this further requires the use of uncommon API (complicated structured dtypes), which is very unlikely to be available to an unprivileged user.
https://github.com/numpy/numpy/issues/18939

Numpy 1.22.0 includes a fix for CVE-2021-41496: Buffer overflow in the array_from_pyobj function of fortranobject.c, which allows attackers to conduct a Denial of Service attacks by carefully constructing an array with negative values. 
NOTE: The vendor does not agree this is a vulnerability; the negative dimensions can only be created by an already privileged user (or internally).
https://github.com/numpy/numpy/issues/19000

Numpy 1.22.2  includes a fix for CVE-2021-41495: Null Pointer Dereference vulnerability exists in numpy.sort in NumPy in the PyArray_DescrNew function due to missing return-value validation, which allows attackers to conduct DoS attacks by repetitively creating sort arrays. 
NOTE: While correct that validation is missing, an error can only occur due to an exhaustion of memory. If the user can exhaust memory, they are already privileged. Further, it should be practically impossible to construct an attack which can target the memory exhaustion to occur at exactly this place.
NOTE2: The specs we include in this advisory differ from the publicly available on other sources. For example, the advisory posted by the NVD indicate that versions up to and including 1.19.0 are affected. However, research by Safety CLI Cybersecurity confirms that the vulnerability remained unaddressed until version 1.22.2.

Pylint before 2.6.1 is susceptible to a Regular Expression Denial of Service (ReDoS) vulnerability due to issues in its pyreverse component. This issue arises from certain regular expressions in pyreverse that can be exploited by causing catastrophic backtracking, significantly slowing down the service by forcing it to take a disproportionate amount of time to process inputs. This vulnerability allows attackers to use specially crafted inputs that increase the processing time exponentially, potentially leading to a service becoming inaccessible to legitimate users.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pylint 2.13.0 fixes a crash when using the doc_params extension.
https://github.com/PyCQA/pylint/issues/5322

Pylint 2.7.0 includes a fix for vulnerable regular expressions in 'pyreverse'.
https://github.com/pylint-dev/pylint/commit/5405dd5115d598fa69e49538d50ec79202b1b52e

Pyyaml version 5.4 includes a fix for CVE-2020-14343: A vulnerability was discovered in the PyYAML library in versions before 5.4, where it is susceptible to arbitrary code execution when it processes untrusted YAML files through the full_load method or with the FullLoader loader. Applications that use the library to process untrusted input may be vulnerable to this flaw. This flaw allows an attacker to execute arbitrary code on the system by abusing the python/object/new constructor. This flaw is due to an incomplete fix for CVE-2020-1747.
https://bugzilla.redhat.com/show_bug.cgi?id=1860466

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in inventory.
https://github.com/sphinx-doc/sphinx/issues/8175
https://github.com/sphinx-doc/sphinx/commit/f7b872e673f9b359a61fd287a7338a28077840d2

Sphinx 3.3.0 includes a fix for a ReDoS vulnerability in docstring.
https://github.com/sphinx-doc/sphinx/issues/8172
https://github.com/sphinx-doc/sphinx/commit/f00e75278c5999f40b214d8934357fbf0e705417

Future 0.18.2 and earlier allows remote attackers to cause a denial of service via crafted Set-Cookie header from malicious web server.
https://github.com/PythonCharmers/python-future/blob/master/src/future/backports/http/cookiejar.py#L215
https://github.com/python/cpython/pull/17157

An oversight in how the Jinja sandboxed environment detects calls to str.format allows an attacker who controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to store a reference to a malicious string's format method, then pass that to a filter that calls it. No such filters are built-in to Jinja, but could be present through custom filters in an application. After the fix, such indirect calls are also handled by the sandbox.

Prior to 3.1.6, an oversight in how the Jinja sandboxed environment interacts with the |attr filter allows an attacker that controls the content of a template to execute arbitrary Python code. To exploit the vulnerability, an attacker needs to control the content of a template. Whether that is the case depends on the type of application using Jinja. This vulnerability impacts users of applications which execute untrusted templates. Jinja's sandbox does catch calls to str.format and ensures they don't escape the sandbox. However, it's possible to use the |attr filter to get a reference to a string's plain format method, bypassing the sandbox. After the fix, the |attr filter no longer bypasses the environment's attribute lookup. This vulnerability is fixed in 3.1.6.

This affects the package jinja2 from 0.0.0 and before 2.11.3. The ReDoS vulnerability is mainly due to the '_punctuation_re regex' operator and its use of multiple wildcards. The last wildcard is the most exploitable as it searches for trailing punctuation. This issue can be mitigated by Markdown to format user content instead of the urlize filter, or by implementing request timeouts and limiting process memory.

Jinja is an extensible templating engine. The `xmlattr` filter in affected versions of Jinja accepts keys containing non-attribute characters. XML/HTML attributes cannot contain spaces, `/`, `>`, or `=`, as each would then be interpreted as starting a separate attribute. If an application accepts keys (as opposed to only values) as user input, and renders these in pages that other users see as well, an attacker could use this to inject other attributes and perform XSS. The fix for CVE-2024-22195 only addressed spaces but not other characters. Accepting keys as user input is now explicitly considered an unintended use case of the `xmlattr` filter, and code that does so without otherwise validating the input should be flagged as insecure, regardless of Jinja version. Accepting _values_ as user input continues to be safe.

Jinja is an extensible templating engine. Special placeholders in the template allow writing code similar to Python syntax. It is possible to inject arbitrary HTML attributes into the rendered HTML template, potentially leading to Cross-Site Scripting (XSS). The Jinja `xmlattr` filter can be abused to inject arbitrary HTML attribute keys and values, bypassing the auto escaping mechanism and potentially leading to XSS. It may also be possible to bypass attribute validation checks if they are blacklist-based.

Ipython versions 8.0.1, 7.31.1, 7.16.3 and 5.11 include a fix for CVE-2022-21699: Affected versions are subject to an arbitrary code execution vulnerability achieved by not properly managing cross user temporary files. This vulnerability allows one user to run code as another on the same machine.
https://github.com/ipython/ipython/security/advisories/GHSA-pq7m-3gw7-gq5x

IPython 8.10.0 includes a fix for CVE-2023-24816: Versions prior to 8.10.0 are subject to a command injection vulnerability with very specific prerequisites. This vulnerability requires that the function 'IPython.utils.terminal.set_term_title' be called on Windows in a Python environment where ctypes is not available. The dependency on 'ctypes' in 'IPython.utils._process_win32' prevents the vulnerable code from ever being reached in the ipython binary. However, as a library that could be used by another tool 'set_term_title' could be called and hence introduce a vulnerability. If an attacker get untrusted input to an instance of this function they would be able to inject shell commands as current process and limited to the scope of the current process. As a workaround, users should ensure that any calls to the 'IPython.utils.terminal.set_term_title' function are done with trusted or filtered input.
https://github.com/ipython/ipython/security/advisories/GHSA-29gw-9793-fvw7

Pygments 2.7.4 includes a fix for CVE-2021-20270: An infinite loop in SMLLexer in Pygments versions 1.5 to 2.7.3 may lead to denial of service when performing syntax highlighting of a Standard ML (SML) source file, as demonstrated by input that only contains the "exception" keyword.

Pygments 2.15.0 includes a fix for CVE-2022-40896: The regular expressions used when parsing Smithy, SQL/SQL+Jinja, and Java properties files were discovered to be vulnerable. As a result, pygmentizing a maliciously-crafted file of these kinds would have resulted in high resources consumption or crashing of the application.
https://pyup.io/posts/pyup-discovers-redos-vulnerabilities-in-top-python-packages-part-2

Pygments 2.7.4 includes a fix for CVE-2021-27291: In pygments 1.1+, fixed in 2.7.4, the lexers used to parse programming languages rely heavily on regular expressions. Some of the regular expressions have exponential or cubic worst-case complexity and are vulnerable to ReDoS. By crafting malicious input, an attacker can cause a denial of service.

Affected versions of Werkzeug are vulnerable to Path Traversal (CWE-22) on Windows systems running Python versions below 3.11. The safe_join() function failed to properly detect certain absolute paths on Windows, allowing attackers to potentially access files outside the intended directory. An attacker could craft special paths starting with "/" that bypass the directory restrictions on Windows systems. The vulnerability exists in the safe_join() function which relied solely on os.path.isabs() for path validation. This is exploitable on Windows systems by passing paths starting with "/" to safe_join(). To remediate, upgrade to the latest version which includes additional path validation checks. 
NOTE: This vulnerability specifically affects Windows systems running Python versions below 3.11 where ntpath.isabs() behavior differs.

Werkzeug 2.2.3 includes a fix for CVE-2023-25577: Prior to version 2.2.3, Werkzeug's multipart form data parser will parse an unlimited number of parts, including file parts. Parts can be a small amount of bytes, but each requires CPU time to parse and may use more memory as Python data. If a request can be made to an endpoint that accesses 'request.data', 'request.form', 'request.files', or 'request.get_data(parse_form_data=False)', it can cause unexpectedly high resource usage. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests. The amount of RAM required can trigger an out of memory kill of the process. Unlimited file parts can use up memory and file handles. If many concurrent requests are sent continuously, this can exhaust or kill all available workers.
https://github.com/pallets/werkzeug/security/advisories/GHSA-xg9f-g7g7-2323

Werkzeug is a comprehensive WSGI web application library. The debugger in affected versions of Werkzeug can allow an attacker to execute code on a developer's machine under some circumstances. This requires the attacker to get the developer to interact with a domain and subdomain they control, and enter the debugger PIN, but if they are successful it allows access to the debugger even if it is only running on localhost. This also requires the attacker to guess a URL in the developer's application that will trigger the debugger.

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in the safe_join function. In Werkzeug versions before 3.1.4, safe_join permits path segments such as “CON” or “AUX” to pass validation, allowing send_from_directory to construct a path that resolves to a Windows device file, which opens successfully but then blocks indefinitely when read.

Werkzeug 2.2.3 includes a fix for CVE-2023-23934: Browsers may allow "nameless" cookies that look like '=value' instead of 'key=value'. A vulnerable browser may allow a compromised application on an adjacent subdomain to exploit this to set a cookie like '=__Host-test=bad' for another subdomain. Werkzeug prior to 2.2.3 will parse the cookie '=__Host-test=bad' as __Host-test=bad'. If a Werkzeug application is running next to a vulnerable or malicious subdomain which sets such a cookie using a vulnerable browser, the Werkzeug application will see the bad cookie value but the valid cookie key.
https://github.com/pallets/werkzeug/security/advisories/GHSA-px8h-6qxv-m22q

Werkzeug is a comprehensive WSGI web application library. If an upload of a file that starts with CR or LF and then is followed by megabytes of data without these characters: all of these bytes are appended chunk by chunk into internal bytearray and lookup for boundary is performed on growing buffer. This allows an attacker to cause a denial of service by sending crafted multipart data to an endpoint that will parse it. The amount of CPU time required can block worker processes from handling legitimate requests.

Affected versions of Werkzeug are potentially vulnerable to resource exhaustion when parsing file data in forms. Applications using 'werkzeug.formparser.MultiPartParser' to parse 'multipart/form-data' requests (e.g. all flask applications) are vulnerable to a relatively simple but effective resource exhaustion (denial of service) attack. A specifically crafted form submission request can cause the parser to allocate and block 3 to 8 times the upload size in main memory. There is no upper limit; a single upload at 1 Gbit/s can exhaust 32 GB of RAM in less than 60 seconds.

Werkzeug 3.0.1 and 2.3.8 include a security fix: Slow multipart parsing for large parts potentially enabling DoS attacks.
https://github.com/pallets/werkzeug/commit/b1916c0c083e0be1c9d887ee2f3d696922bfc5c1

Affected versions of the Werkzeug package are vulnerable to Denial of Service (DoS) due to improper handling of Windows special device names in path joining logic. The safe_join() function fails to reject device-name path segments when they are preceded by other segments (for example, example/NUL), and send_from_directory() relies on safe_join() when resolving user-supplied file paths.

Affected versions of the Werkzeug package are vulnerable to Improper Handling of Windows Device Names due to incomplete validation of Windows reserved device names in user-controlled path segments. In werkzeug.security.safe_join, path components ending with special device names (for example CON or AUX) are not reliably rejected when they include compound extensions (for example CON.txt.html) or trailing spaces, and werkzeug.utils.send_from_directory relies on safe_join when resolving a requested filename.

Affected versions of the filelock package are vulnerable to a Time-of-Check Time-of-Use (TOCTOU) Race Condition due to a race window between a write-permission check and lock file creation that does not prevent symlink substitution. The flaw is in filelock.SoftFileLock in src/filelock/_soft.py, where _acquire() calls raise_on_not_writable_file() and then performs os.open() on the lock path, allowing the filesystem state to change between the check and the use.

Affected versions of this package are vulnerable to Time-of-Check Time-of-Use (TOCTOU) Race Condition. The file locking mechanism in SoftFileLock._acquire() performs permission validation before file creation without using the O_NOFOLLOW flag, leading to a race window where attackers with local access can create symlinks that redirect lock operations to arbitrary files. An attacker can exploit this vulnerability by creating a malicious symlink during the brief window between permission check and file creation, causing the lock to operate on unintended target files and potentially enabling unauthorized access or file corruption.

Affected versions of this package are vulnerable to a TOCTOU (Time-of-Check to Time-of-Use) symlink vulnerability due to improper handling of symlinks during lock file creation. The vulnerability exists because the package does not adequately check for symlink manipulation between the time the lock file path is checked and the time it is used. An attacker can exploit this vulnerability by creating a malicious symlink, potentially leading to unauthorized access or modification of files, which could compromise the integrity and security of the system.

Affected versions of the gunicorn package are vulnerable to HTTP Request/Response Smuggling due to improper validation of the Transfer-Encoding header that enables a TE.CL desynchronisation condition. Gunicorn’s HTTP parser does not consistently enforce RFC semantics when parsing conflicting or unsupported request framing—such as messages containing both Transfer-Encoding and Content-Length—so the backend may fall back to Content-Length while an intermediary treats the message as chunked.

Gunicorn fails to properly validate Transfer-Encoding headers, leading to HTTP Request Smuggling (HRS) vulnerabilities. By crafting requests with conflicting Transfer-Encoding headers, attackers can bypass security restrictions and access restricted endpoints. This issue is due to Gunicorn's handling of Transfer-Encoding headers, where it incorrectly processes requests with multiple, conflicting Transfer-Encoding headers, treating them as chunked regardless of the final encoding specified. This vulnerability allows for a range of attacks including cache poisoning, session manipulation, and data exposure.

Affected versions of this package are potentially vulnerable to Regular Expression Denial of Service (ReDoS) due to catastrophic backtracking in the V1 engine when processing patterns that combine full‑casefolding with the [\s\S]* quantifier. The engine’s AnyAll node fails to prevent nested quantifier backtracking, leading to infinite loops and CPU exhaustion.

Affected versions of Requests, when making requests through a Requests `Session`, if the first request is made with `verify=False` to disable cert verification, all subsequent requests to the same host will continue to ignore cert verification regardless of changes to the value of `verify`. This behavior will continue for the lifecycle of the connection in the connection pool. Requests 2.32.0 fixes the issue, but versions 2.32.0 and 2.32.1 were yanked due to conflicts with CVE-2024-35195 mitigation.

Affected versions of the requests package are vulnerable to Insecure Temporary File reuse due to predictable temporary filename generation in extract_zipped_paths(). The requests.utils.extract_zipped_paths() utility extracts files from zip archives into the system temporary directory using a deterministic path, and if that file already exists, the function reuses it without validating that it is the expected extracted content.

Affected versions of Requests are vulnerable to proxy credential leakage. When redirected to an HTTPS endpoint, the Proxy-Authorization header is forwarded to the destination server due to the use of rebuild_proxies to reattach the header. This may allow a malicious actor to exfiltrate sensitive information.

Requests is an HTTP library. Due to a URL parsing issue, Requests releases prior to 2.32.4 may leak .netrc credentials to third parties for specific maliciously-crafted URLs. Users should upgrade to version 2.32.4 to receive a fix. For older versions of Requests, use of the .netrc file can be disabled with `trust_env=False` on one's Requests Session.

Urllib3's ProxyManager ensures that the Proxy-Authorization header is correctly directed only to configured proxies. However, when HTTP requests bypass urllib3's proxy support, there's a risk of inadvertently setting the Proxy-Authorization header, which remains ineffective without a forwarding or tunneling proxy. Urllib3 does not recognize this header as carrying authentication data, failing to remove it during cross-origin redirects. While this scenario is uncommon and poses low risk to most users, urllib3 now proactively removes the Proxy-Authorization header during cross-origin redirects as a precautionary measure. Users are advised to utilize urllib3's proxy support or disable automatic redirects to handle the Proxy-Authorization header securely. Despite these precautions, urllib3 defaults to stripping the header to safeguard users who may inadvertently misconfigure requests.

Affected versions of the urllib3 package are vulnerable to Information Disclosure due to improper handling of sensitive headers during cross-origin redirects in the low-level proxy API. When following cross-origin redirects via ProxyManager.connection_from_url().urlopen() with assert_same_host=False, sensitive headers including Authorization, Cookie, and Proxy-Authorization are not stripped, unlike the high-level API, which removes them via Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT. An attacker controlling a redirect target can capture these sensitive headers from requests that follow cross-origin redirects through the low-level proxy API path.

urllib3 is a user-friendly HTTP client library for Python. Prior to 2.5.0, it is possible to disable redirects for all requests by instantiating a PoolManager and specifying retries in a way that disable redirects. By default, requests and botocore users are not affected. An application attempting to mitigate SSRF or open redirect vulnerabilities by disabling redirects at the PoolManager level will remain vulnerable. This issue has been patched in version 2.5.0.

Affected versions of urllib3 are vulnerable to an HTTP redirect handling vulnerability that fails to remove the HTTP request body when a POST changes to a GET via 301, 302, or 303 responses. This flaw can expose sensitive request data if the origin service is compromised and redirects to a malicious endpoint, though exploitability is low when no sensitive data is used. The vulnerability affects automatic redirect behavior. It is fixed in versions 1.26.18 and 2.0.7; update or disable redirects using redirects=False.
This vulnerability is specific to Python's urllib3 library.

Urllib3 1.26.17 and 2.0.5 include a fix for CVE-2023-43804: Urllib3 doesn't treat the 'Cookie' HTTP header special or provide any helpers for managing cookies over HTTP, that is the responsibility of the user. However, it is possible for a user to specify a 'Cookie' header and unknowingly leak information via HTTP redirects to a different origin if that user doesn't disable redirects explicitly.
https://github.com/urllib3/urllib3/security/advisories/GHSA-v845-jxx5-vc9f

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to improper handling of highly compressed HTTP response bodies during streaming decompression. The urllib3.HTTPResponse methods stream(), read(), read1(), read_chunked(), and readinto() may fully decompress a minimal but highly compressed payload based on the Content-Encoding header into an internal buffer instead of limiting the decompressed output to the requested chunk size, causing excessive CPU usage and massive memory allocation on the client side.

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to redirect handling that drains connections by decompressing redirect response bodies without enforcing streaming read limits. The issue occurs when using urllib3’s streaming mode (for example, preload_content=False) while allowing redirects, because urllib3.response.HTTPResponse.drain_conn() would call HTTPResponse.read() in a way that decoded/decompressed the entire redirect response body even before any streaming reads were performed, effectively bypassing decompression-bomb safeguards.

Urllib3 1.26.5 includes a fix for CVE-2021-33503: When provided with a URL containing many @ characters in the authority component, the authority regular expression exhibits catastrophic backtracking, causing a denial of service if a URL were passed as a parameter or redirected to via an HTTP redirect.
https://github.com/advisories/GHSA-q2q7-5pp4-w6pg

Affected versions of the urllib3 package are vulnerable to Denial of Service (DoS) due to allowing an unbounded number of content-encoding decompression steps for HTTP responses. The HTTPResponse content decoding pipeline in urllib3 follows the Content-Encoding header and applies each advertised compression algorithm in sequence without enforcing a maximum chain length or effective output size, so a malicious peer can send a response with a very long encoding chain that triggers excessive CPU use and massive memory allocation during decompression.

Selenium 4.15.1 (Python bindings) include a fix for CVE-2023-5590: NULL Pointer Dereference.
https://github.com/seleniumhq/selenium/commit/023a0d52f106321838ab1c0997e76693f4dcbdf6

Affected versions of the virtualenv package (up to and including 20.36.1) are vulnerable to Race Condition (TOCTOU) attacks due to non-atomic directory creation that is performed using check-then-act filesystem logic. The issue occurs in virtualenv’s directory creation operations for its app_data path and related lock file handling, where a directory existence check can be raced so a symlink is inserted before the subsequent creation or access step, redirecting operations to an unintended location.

Virtualenv version 20.21.0 addresses a race condition in `virtualenv.cli_run` where a `FileNotFoundError` could occur for a JSON file in `pypa/virtualenv/py_info/1`. This error happens if the underlying interpreter is updated, causing the JSON file to be deleted and rewritten.

Affected versions of the virtualenv package are vulnerable to Command Injection due to improper quoting of template string placeholders in activation scripts. The vulnerability exists in the ViaTemplateActivator class, where magic template strings like __VIRTUAL_ENV__ are replaced in shell activation scripts without proper escaping or quoting, allowing shell metacharacters to be interpreted as commands during string substitution. An attacker can exploit this vulnerability by creating a virtual environment with a specially crafted directory name containing shell commands (such as "';uname -a;':"), which will be executed when the activation script is sourced, resulting in arbitrary command execution with the privileges of the user activating the virtual environment.

Prompt-toolkit 3.0.13 fixes a race condition in `ThreadedHistory` which could lead to a deadlock. 
https://github.com/prompt-toolkit/python-prompt-toolkit/commit/99092a8c6d4b411645ac4b84d504e5226e7eebb8#diff-48c0ff10dc3990285d19b3f54e6bfec763089ba1229dc6f9e88463a1046adad7R163

Dash-html-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.

Dash-core-components affected versions are vulnerable to Cross-site Scripting (XSS) when the href of the a tag is controlled by an adversary. An authenticated attacker who stores a view that exploits this vulnerability could steal the data that's visible to another user who opens that view - not just the data already included on the page, but they could also, in theory, make additional requests and access other data accessible to this user. In some cases, they could also steal the access tokens of that user, which would allow the attacker to act as that user, including viewing other apps and resources hosted on the same server. 
#Note: This is only exploitable in Dash apps that include some mechanism to store user input to be reloaded by a different user.