Vulnerability-Lookup

URL	Tags
https://github.com/Kludex/python-multipart/securi…	x_refsource_CONFIRM

Vendor	Product	Version	CPE status
Kludex	python-multipart	Affected: < 0.0.30	guessed

BREW-CYCODE-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:42 – Updated: 2026-09-17 18:32 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "cycode",
        "purl": "pkg:brew/cycode"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "3.2.0"
            },
            {
              "fixed": "3.16.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-cycode-CVE-2026-53538",
  "modified": "2026-09-17T18:32:20Z",
  "published": "2026-08-13T16:42:07Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-DATASETTE-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:42 – Updated: 2026-09-17 17:44 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "datasette",
        "purl": "pkg:brew/datasette"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.47.3"
            },
            {
              "fixed": "0.65.2_5"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-datasette-CVE-2026-53538",
  "modified": "2026-09-17T17:44:18Z",
  "published": "2026-08-13T16:42:09Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-DSTACK-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:43 – Updated: 2026-09-17 18:14 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "dstack",
        "purl": "pkg:brew/dstack"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.12.1"
            },
            {
              "fixed": "0.20.23"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-dstack-CVE-2026-53538",
  "modified": "2026-09-17T18:14:18Z",
  "published": "2026-08-13T16:43:57Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-FASTAPI-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:46 – Updated: 2026-09-17 19:08 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "fastapi",
        "purl": "pkg:brew/fastapi"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.111.0"
            },
            {
              "fixed": "0.137.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-fastapi-CVE-2026-53538",
  "modified": "2026-09-17T19:08:32Z",
  "published": "2026-08-13T16:46:30Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-FASTMCP-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:46 – Updated: 2026-09-17 19:27 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "fastmcp",
        "purl": "pkg:brew/fastmcp"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "2.12.3"
            },
            {
              "fixed": "3.4.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-fastmcp-CVE-2026-53538",
  "modified": "2026-09-17T19:27:54Z",
  "published": "2026-08-13T16:46:31Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-GPTME-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:54 – Updated: 2026-09-17 18:12 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "gptme",
        "purl": "pkg:brew/gptme"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.28.0"
            },
            {
              "fixed": "0.31.0_12"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-gptme-CVE-2026-53538",
  "modified": "2026-09-17T18:12:06Z",
  "published": "2026-08-13T16:54:46Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-HERMES-AGENT-CVE-20… (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 16:56 – Updated: 2026-09-10 19:27 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "hermes-agent",
        "purl": "pkg:brew/hermes-agent"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2026.8.3_1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-hermes-agent-CVE-2026-53538",
  "modified": "2026-09-10T19:27:17Z",
  "published": "2026-08-13T16:56:26Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-KIMI-CLI-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 17:01 – Updated: 2026-09-17 20:03 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "kimi-cli",
        "purl": "pkg:brew/kimi-cli"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.40"
            },
            {
              "fixed": "1.47.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-kimi-cli-CVE-2026-53538",
  "modified": "2026-09-17T20:03:17Z",
  "published": "2026-08-13T17:01:40Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-MCP-ATLASSIAN-CVE-2… (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 17:13 – Updated: 2026-09-17 19:03 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "mcp-atlassian",
        "purl": "pkg:brew/mcp-atlassian"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.11.9"
            },
            {
              "fixed": "0.21.1_8"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-mcp-atlassian-CVE-2026-53538",
  "modified": "2026-09-17T19:03:07Z",
  "published": "2026-08-13T17:13:13Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

BREW-MCP-GOOGLE-SHEETS-C… (GHSA-6JV3-5F52-599M)

Vulnerability from osv_homebrew – Published: 2026-08-13 17:13 – Updated: 2026-09-17 18:01 – Source website

Summary

python-multipart: Semicolon treated as querystring field separator enables parameter smuggling

Details

Summary

QuerystringParser treated ; as a field separator in application/x-www-form-urlencoded bodies, in addition to &. The WHATWG URL standard, modern browsers, and Python's urllib.parse (since the CVE-2021-23336 fix) treat only & as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.

Details

In python_multipart/multipart.py, the FIELD_NAME and FIELD_DATA states located the next separator by scanning for & and, failing that, for ;:

sep_pos = data.find(b"&", i)
if sep_pos == -1:
    sep_pos = data.find(b";", i)

As a result, ; acted as a field boundary. Because the fallback only triggered when no & remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across write() calls. This is the same class of issue as CVE-2021-23336 in CPython's urllib.parse.

For example, a body inspecting WAF or gateway that follows the WHATWG rule (only & separates fields) receives:

role=user&x=;role=admin

The upstream parses two fields, role=user and x=";role=admin", sees a benign role=user, and forwards the request. QuerystringParser parsed the same bytes as three fields: role="user", x="", and role="admin". The application (for example via Starlette/FastAPI request.form(), where the last value wins) then received role=admin, a value the upstream validator never saw.

The parser is reachable through the public QuerystringParser class, the high level FormParser, create_form_parser, and parse_form APIs, and Starlette/FastAPI request.form() for url encoded bodies.

Impact

Interpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.

Mitigation

Upgrade to python-multipart 0.0.30 or later, which treats only & as a field separator per the WHATWG URL standard. ; is parsed as ordinary field data, matching urllib.parse, browsers, and other compliant parsers.

Severity

3.7 (Low)


                    
                      CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N

References

URL

Type

	https://github.com/Kludex/python-multipart/securi…	WEB
	https://github.com/Kludex/python-multipart	PACKAGE

JSON

To clipboard

{
  "affected": [
    {
      "ecosystem_specific": {
        "fix": "bump",
        "range_state": "fixed",
        "resource": "python-multipart",
        "resource_purl": "pkg:pypi/python-multipart@0.0.32",
        "upstream_fixed_in": "0.0.30"
      },
      "package": {
        "ecosystem": "Homebrew",
        "name": "mcp-google-sheets",
        "purl": "pkg:brew/mcp-google-sheets"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0.5.1"
            },
            {
              "fixed": "0.6.3_2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "database_specific": {
    "confidence": "high",
    "source": "matched",
    "strategy": "registry",
    "upstream_evidence": [
      {
        "ecosystem": "PyPI",
        "key": "pkg:pypi/python-multipart@0.0.32",
        "name": "python-multipart",
        "resource": "python-multipart",
        "strategy": "registry",
        "subject_version": "0.0.32"
      }
    ]
  },
  "details": "### Summary\n\n`QuerystringParser` treated `;` as a field separator in `application/x-www-form-urlencoded` bodies, in addition to `\u0026`. The [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing), modern browsers, and Python\u0027s `urllib.parse` (since the CVE-2021-23336 fix) treat only `\u0026` as a separator. This creates a parser differential: the same bytes are tokenized into different fields than a WHATWG compliant intermediary would produce, allowing an attacker to smuggle extra form fields past an upstream body inspecting component.\n\n### Details\n\nIn `python_multipart/multipart.py`, the `FIELD_NAME` and `FIELD_DATA` states located the next separator by scanning for `\u0026` and, failing that, for `;`:\n\n```python\nsep_pos = data.find(b\"\u0026\", i)\nif sep_pos == -1:\n    sep_pos = data.find(b\";\", i)\n```\n\nAs a result, `;` acted as a field boundary. Because the fallback only triggered when no `\u0026` remained in the current chunk, tokenization also depended on unrelated bytes later in the buffer and on how the body was split across `write()` calls. This is the same class of issue as CVE-2021-23336 in CPython\u0027s `urllib.parse`.\n\nFor example, a body inspecting WAF or gateway that follows the WHATWG rule (only `\u0026` separates fields) receives:\n\n```\nrole=user\u0026x=;role=admin\n```\n\nThe upstream parses two fields, `role=user` and `x=\";role=admin\"`, sees a benign `role=user`, and forwards the request. `QuerystringParser` parsed the same bytes as three fields: `role=\"user\"`, `x=\"\"`, and `role=\"admin\"`. The application (for example via Starlette/FastAPI `request.form()`, where the last value wins) then received `role=admin`, a value the upstream validator never saw.\n\nThe parser is reachable through the public `QuerystringParser` class, the high level `FormParser`, `create_form_parser`, and `parse_form` APIs, and Starlette/FastAPI `request.form()` for url encoded bodies.\n\n### Impact\n\nInterpretation conflict / HTTP parameter pollution. An attacker can smuggle extra or overriding form fields past an upstream component that applies the WHATWG separator rule, reaching the backend with parameters the intermediary did not observe.\n\n### Mitigation\n\nUpgrade to `python-multipart` `0.0.30` or later, which treats only `\u0026` as a field separator per the [WHATWG URL standard](https://url.spec.whatwg.org/#urlencoded-parsing). `;` is parsed as ordinary field data, matching `urllib.parse`, browsers, and other compliant parsers.",
  "id": "BREW-mcp-google-sheets-CVE-2026-53538",
  "modified": "2026-09-17T18:01:45Z",
  "published": "2026-08-13T17:13:13Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-6jv3-5f52-599m"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Kludex/python-multipart"
    }
  ],
  "schema_version": "1.7.3",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "python-multipart: Semicolon treated as querystring field separator enables parameter smuggling",
  "upstream": [
    "GHSA-6jv3-5f52-599m",
    "CVE-2026-53538",
    "PYSEC-2026-3037"
  ]
}

CVE-2026-53538 (GCVE-0-2026-53538)

Tags

Sightings

Nomenclature

Related by attack behaviour

Action not permitted

CVE-2026-53538 (GCVE-0-2026-53538)

BREW-CYCODE-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-DATASETTE-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-DSTACK-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-FASTAPI-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-FASTMCP-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-GPTME-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-HERMES-AGENT-CVE-20… (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-KIMI-CLI-CVE-2026-53538 (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-MCP-ATLASSIAN-CVE-2… (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

BREW-MCP-GOOGLE-SHEETS-C… (GHSA-6JV3-5F52-599M)

Summary

Details

Impact

Mitigation

Tags

Sightings

Nomenclature

Related by attack behaviour