Vulnerability-Lookup

CVE-2026-25640 (GCVE-0-2026-25640)

Vulnerability from cvelistv5 – Published: 2026-02-06 20:01 – Updated: 2026-06-30 12:06

Title

Pydantic AI affected by Stored XSS via Path Traversal in Web UI CDN URL

Summary

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.34.0 to before 1.51.0, a path traversal vulnerability in the Pydantic AI web UI allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. In affected versions, the CDN URL is constructed using a version query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data. This vulnerability only affects applications that use Agent.to_web to serve a chat interface and clai web to serve a chat interface from the CLI. These are typically run locally (on localhost), but may also be deployed on a remote server. This vulnerability is fixed in 1.51.0.

Severity

7.1 (High)


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

SSVC

Exploitation: none Automatable: no Technical Impact: partial

CISA Coordinator (v2.0.3)

CWE

CWE-22 - Improper Limitation of a Pathname to a Restricted Directory ('Path Traversal')
CWE-79 - Improper Neutralization of Input During Web Page Generation ('Cross-site Scripting')

Assigner

GitHub_M

References

5 references

URL	Tags
https://github.com/pydantic/pydantic-ai/security/…	x_refsource_CONFIRM
https://github.com/pydantic/pydantic-ai/releases/…	x_refsource_MISC
https://access.redhat.com/security/cve/CVE-2026-25640	vdb-entryx_refsource_REDHAT
https://bugzilla.redhat.com/show_bug.cgi?id=2437753	issue-trackingx_refsource_REDHAT
https://security.access.redhat.com/data/csaf/v2/v…	x_sadp-csaf-vex

Impacted products

2 products

Vendor	Product	Version
pydantic	pydantic-ai	Affected: >= 1.34.0, < 1.51.0
Red Hat	Red Hat Enterprise Linux AI (RHEL AI) 3	cpe:/a:redhat:enterprise_linux_ai:3

Show details on NVD website

JSON

To clipboard

{
  "containers": {
    "adp": [
      {
        "metrics": [
          {
            "other": {
              "content": {
                "id": "CVE-2026-25640",
                "options": [
                  {
                    "Exploitation": "none"
                  },
                  {
                    "Automatable": "no"
                  },
                  {
                    "Technical Impact": "partial"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2026-02-06T20:16:17.763231Z",
                "version": "2.0.3"
              },
              "type": "ssvc"
            }
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2026-02-06T20:16:23.201Z",
          "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
          "shortName": "CISA-ADP"
        },
        "title": "CISA ADP Vulnrichment"
      },
      {
        "affected": [
          {
            "cpes": [
              "cpe:/a:redhat:enterprise_linux_ai:3"
            ],
            "defaultStatus": "affected",
            "product": "Red Hat Enterprise Linux AI (RHEL AI) 3",
            "vendor": "Red Hat"
          }
        ],
        "datePublic": "2026-02-06T20:01:53.892Z",
        "descriptions": [
          {
            "lang": "en",
            "value": "A flaw was found in Pydantic AI. A remote attacker can exploit a path traversal vulnerability in the Pydantic AI web UI by crafting a malicious URL. This vulnerability arises from insufficient validation of the version query parameter, allowing the server to fetch and serve attacker-controlled HTML or JavaScript. If a victim clicks on such a link, malicious code can execute in their browser, potentially leading to the theft of sensitive client-side data, including chat history."
          }
        ],
        "metrics": [
          {
            "other": {
              "content": {
                "namespace": "https://access.redhat.com/security/updates/classification/",
                "value": "Important"
              },
              "type": "Red Hat severity rating"
            }
          },
          {
            "cvssV3_1": {
              "attackComplexity": "LOW",
              "attackVector": "NETWORK",
              "availabilityImpact": "NONE",
              "baseScore": 7.1,
              "baseSeverity": "HIGH",
              "confidentialityImpact": "HIGH",
              "integrityImpact": "LOW",
              "privilegesRequired": "NONE",
              "scope": "UNCHANGED",
              "userInteraction": "REQUIRED",
              "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N",
              "version": "3.1"
            },
            "format": "CVSS"
          }
        ],
        "problemTypes": [
          {
            "descriptions": [
              {
                "cweId": "CWE-22",
                "description": "Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)",
                "lang": "en",
                "type": "CWE"
              }
            ]
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2026-06-30T12:06:32.866Z",
          "orgId": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c",
          "shortName": "redhat-SADP"
        },
        "references": [
          {
            "tags": [
              "vdb-entry",
              "x_refsource_REDHAT"
            ],
            "url": "https://access.redhat.com/security/cve/CVE-2026-25640"
          },
          {
            "name": "RHBZ#2437753",
            "tags": [
              "issue-tracking",
              "x_refsource_REDHAT"
            ],
            "url": "https://bugzilla.redhat.com/show_bug.cgi?id=2437753"
          },
          {
            "tags": [
              "x_sadp-csaf-vex"
            ],
            "url": "https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-25640.json"
          }
        ],
        "timeline": [
          {
            "lang": "en",
            "time": "2026-02-09T11:03:46.831Z",
            "value": "Reported to Red Hat."
          },
          {
            "lang": "en",
            "time": "2026-02-06T20:01:53.892Z",
            "value": "Made public."
          }
        ],
        "title": "Pydantic AI: Pydantic AI: Arbitrary code execution and information disclosure via path traversal in web UI",
        "workarounds": [
          {
            "lang": "en",
            "value": "To reduce exposure, restrict network access to the Pydantic AI web interface. If the application is deployed on a remote server, configure firewall rules to limit access to trusted networks or localhost. This prevents untrusted clients from reaching the vulnerable component. After applying firewall changes, ensure the rules are persistent and the firewall service is reloaded."
          }
        ],
        "x_adpType": "supplier",
        "x_generator": {
          "engine": "sadp-cli 1.0.0"
        }
      }
    ],
    "cna": {
      "affected": [
        {
          "product": "pydantic-ai",
          "vendor": "pydantic",
          "versions": [
            {
              "status": "affected",
              "version": "\u003e= 1.34.0, \u003c 1.51.0"
            }
          ]
        }
      ],
      "descriptions": [
        {
          "lang": "en",
          "value": "Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.34.0 to before 1.51.0,  a path traversal vulnerability in the Pydantic AI web UI allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. In affected versions, the CDN URL is constructed using a version query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data. This vulnerability only affects applications that use Agent.to_web to serve a chat interface and clai web to serve a chat interface from the CLI. These are typically run locally (on localhost), but may also be deployed on a remote server. This vulnerability is fixed in 1.51.0."
        }
      ],
      "metrics": [
        {
          "cvssV3_1": {
            "attackComplexity": "LOW",
            "attackVector": "NETWORK",
            "availabilityImpact": "NONE",
            "baseScore": 7.1,
            "baseSeverity": "HIGH",
            "confidentialityImpact": "HIGH",
            "integrityImpact": "LOW",
            "privilegesRequired": "NONE",
            "scope": "UNCHANGED",
            "userInteraction": "REQUIRED",
            "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N",
            "version": "3.1"
          }
        }
      ],
      "problemTypes": [
        {
          "descriptions": [
            {
              "cweId": "CWE-22",
              "description": "CWE-22: Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)",
              "lang": "en",
              "type": "CWE"
            }
          ]
        },
        {
          "descriptions": [
            {
              "cweId": "CWE-79",
              "description": "CWE-79: Improper Neutralization of Input During Web Page Generation (\u0027Cross-site Scripting\u0027)",
              "lang": "en",
              "type": "CWE"
            }
          ]
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2026-02-06T20:01:53.892Z",
        "orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
        "shortName": "GitHub_M"
      },
      "references": [
        {
          "name": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7",
          "tags": [
            "x_refsource_CONFIRM"
          ],
          "url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7"
        },
        {
          "name": "https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0",
          "tags": [
            "x_refsource_MISC"
          ],
          "url": "https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0"
        }
      ],
      "source": {
        "advisory": "GHSA-wjp5-868j-wqv7",
        "discovery": "UNKNOWN"
      },
      "title": "Pydantic AI affected by Stored XSS via Path Traversal in Web UI CDN URL"
    }
  },
  "cveMetadata": {
    "assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
    "assignerShortName": "GitHub_M",
    "cveId": "CVE-2026-25640",
    "datePublished": "2026-02-06T20:01:53.892Z",
    "dateReserved": "2026-02-04T05:15:41.791Z",
    "dateUpdated": "2026-06-30T12:06:32.866Z",
    "state": "PUBLISHED"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.2",
  "vulnerability-lookup:meta": {
    "epss": {
      "cve": "CVE-2026-25640",
      "date": "2026-07-05",
      "epss": "0.00324",
      "percentile": "0.2425"
    },
    "nvd": "{\"cve\":{\"id\":\"CVE-2026-25640\",\"sourceIdentifier\":\"security-advisories@github.com\",\"published\":\"2026-02-06T20:16:11.110\",\"lastModified\":\"2026-06-30T03:17:43.313\",\"vulnStatus\":\"Modified\",\"cveTags\":[],\"descriptions\":[{\"lang\":\"en\",\"value\":\"Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.34.0 to before 1.51.0,  a path traversal vulnerability in the Pydantic AI web UI allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. In affected versions, the CDN URL is constructed using a version query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data. This vulnerability only affects applications that use Agent.to_web to serve a chat interface and clai web to serve a chat interface from the CLI. These are typically run locally (on localhost), but may also be deployed on a remote server. This vulnerability is fixed in 1.51.0.\"},{\"lang\":\"es\",\"value\":\"Pydantic AI es un framework de agente Python para construir aplicaciones y flujos de trabajo con IA Generativa. Desde la 1.34.0 hasta antes de la 1.51.0, una vulnerabilidad de salto de ruta en la interfaz de usuario web de Pydantic AI permite a un atacante servir JavaScript arbitrario en el contexto de la aplicaci\u00f3n al crear una URL maliciosa. En las versiones afectadas, la URL de la CDN se construye utilizando un par\u00e1metro de consulta de versi\u00f3n de la URL de la solicitud. Este par\u00e1metro no se valida, lo que permite secuencias de salto de ruta que hacen que el servidor obtenga y sirva HTML/JavaScript controlado por el atacante desde una fuente arbitraria en la misma CDN, en lugar del paquete leg\u00edtimo de la interfaz de usuario de chat. Si una v\u00edctima hace clic en el enlace o lo visita a trav\u00e9s de un iframe, el c\u00f3digo controlado por el atacante se ejecuta en su navegador, lo que permite el robo del historial de chat y otros datos del lado del cliente. Esta vulnerabilidad solo afecta a las aplicaciones que usan Agent.to_web para servir una interfaz de chat y clai web para servir una interfaz de chat desde la CLI. Estas suelen ejecutarse localmente (en localhost), pero tambi\u00e9n pueden implementarse en un servidor remoto. Esta vulnerabilidad se corrige en la 1.51.0.\"}],\"affected\":[{\"source\":\"security-advisories@github.com\",\"affectedData\":[{\"vendor\":\"pydantic\",\"product\":\"pydantic-ai\",\"versions\":[{\"version\":\"\u003e= 1.34.0, \u003c 1.51.0\",\"status\":\"affected\"}]}]},{\"source\":\"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\",\"affectedData\":[{\"vendor\":\"Red Hat\",\"product\":\"Red Hat Enterprise Linux AI (RHEL AI) 3\",\"defaultStatus\":\"affected\",\"cpes\":[\"cpe:/a:redhat:enterprise_linux_ai:3\"]}]}],\"metrics\":{\"cvssMetricV31\":[{\"source\":\"security-advisories@github.com\",\"type\":\"Secondary\",\"cvssData\":{\"version\":\"3.1\",\"vectorString\":\"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N\",\"baseScore\":7.1,\"baseSeverity\":\"HIGH\",\"attackVector\":\"NETWORK\",\"attackComplexity\":\"LOW\",\"privilegesRequired\":\"NONE\",\"userInteraction\":\"REQUIRED\",\"scope\":\"UNCHANGED\",\"confidentialityImpact\":\"HIGH\",\"integrityImpact\":\"LOW\",\"availabilityImpact\":\"NONE\"},\"exploitabilityScore\":2.8,\"impactScore\":4.2},{\"source\":\"nvd@nist.gov\",\"type\":\"Primary\",\"cvssData\":{\"version\":\"3.1\",\"vectorString\":\"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:L/A:N\",\"baseScore\":5.4,\"baseSeverity\":\"MEDIUM\",\"attackVector\":\"NETWORK\",\"attackComplexity\":\"LOW\",\"privilegesRequired\":\"NONE\",\"userInteraction\":\"REQUIRED\",\"scope\":\"UNCHANGED\",\"confidentialityImpact\":\"LOW\",\"integrityImpact\":\"LOW\",\"availabilityImpact\":\"NONE\"},\"exploitabilityScore\":2.8,\"impactScore\":2.5},{\"source\":\"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\",\"type\":\"Secondary\",\"cvssData\":{\"version\":\"3.1\",\"vectorString\":\"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N\",\"baseScore\":7.1,\"baseSeverity\":\"HIGH\",\"attackVector\":\"NETWORK\",\"attackComplexity\":\"LOW\",\"privilegesRequired\":\"NONE\",\"userInteraction\":\"REQUIRED\",\"scope\":\"UNCHANGED\",\"confidentialityImpact\":\"HIGH\",\"integrityImpact\":\"LOW\",\"availabilityImpact\":\"NONE\"},\"exploitabilityScore\":2.8,\"impactScore\":4.2}],\"ssvcV203\":[{\"source\":\"134c704f-9b21-4f2e-91b3-4a467353bcc0\",\"ssvcData\":{\"timestamp\":\"2026-02-06T20:16:17.763231Z\",\"id\":\"CVE-2026-25640\",\"options\":[{\"exploitation\":\"none\"},{\"automatable\":\"no\"},{\"technicalImpact\":\"partial\"}],\"role\":\"CISA Coordinator\",\"version\":\"2.0.3\"}}]},\"weaknesses\":[{\"source\":\"security-advisories@github.com\",\"type\":\"Primary\",\"description\":[{\"lang\":\"en\",\"value\":\"CWE-22\"},{\"lang\":\"en\",\"value\":\"CWE-79\"}]},{\"source\":\"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\",\"type\":\"Secondary\",\"description\":[{\"lang\":\"en\",\"value\":\"CWE-22\"}]}],\"configurations\":[{\"nodes\":[{\"operator\":\"OR\",\"negate\":false,\"cpeMatch\":[{\"vulnerable\":true,\"criteria\":\"cpe:2.3:a:pydantic:pydantic_ai:*:*:*:*:*:python:*:*\",\"versionStartIncluding\":\"1.34.0\",\"versionEndExcluding\":\"1.51.0\",\"matchCriteriaId\":\"CD7C18E2-854E-44FD-A4AF-D4E580B67D98\"}]}]}],\"references\":[{\"url\":\"https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0\",\"source\":\"security-advisories@github.com\",\"tags\":[\"Product\",\"Release Notes\"]},{\"url\":\"https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7\",\"source\":\"security-advisories@github.com\",\"tags\":[\"Vendor Advisory\"]},{\"url\":\"https://access.redhat.com/security/cve/CVE-2026-25640\",\"source\":\"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\"},{\"url\":\"https://bugzilla.redhat.com/show_bug.cgi?id=2437753\",\"source\":\"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\"},{\"url\":\"https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-25640.json\",\"source\":\"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\"}]}}",
    "vulnrichment": {
      "containers": "{\"adp\": [{\"title\": \"Pydantic AI: Pydantic AI: Arbitrary code execution and information disclosure via path traversal in web UI\", \"metrics\": [{\"other\": {\"type\": \"Red Hat severity rating\", \"content\": {\"value\": \"Important\", \"namespace\": \"https://access.redhat.com/security/updates/classification/\"}}}, {\"format\": \"CVSS\", \"cvssV3_1\": {\"scope\": \"UNCHANGED\", \"version\": \"3.1\", \"baseScore\": 7.1, \"attackVector\": \"NETWORK\", \"baseSeverity\": \"HIGH\", \"vectorString\": \"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N\", \"integrityImpact\": \"LOW\", \"userInteraction\": \"REQUIRED\", \"attackComplexity\": \"LOW\", \"availabilityImpact\": \"NONE\", \"privilegesRequired\": \"NONE\", \"confidentialityImpact\": \"HIGH\"}}], \"affected\": [{\"cpes\": [\"cpe:/a:redhat:enterprise_linux_ai:3\"], \"vendor\": \"Red Hat\", \"product\": \"Red Hat Enterprise Linux AI (RHEL AI) 3\", \"defaultStatus\": \"affected\"}], \"timeline\": [{\"lang\": \"en\", \"time\": \"2026-02-09T11:03:46.831Z\", \"value\": \"Reported to Red Hat.\"}, {\"lang\": \"en\", \"time\": \"2026-02-06T20:01:53.892Z\", \"value\": \"Made public.\"}], \"x_adpType\": \"supplier\", \"datePublic\": \"2026-02-06T20:01:53.892Z\", \"references\": [{\"url\": \"https://access.redhat.com/security/cve/CVE-2026-25640\", \"tags\": [\"vdb-entry\", \"x_refsource_REDHAT\"]}, {\"url\": \"https://bugzilla.redhat.com/show_bug.cgi?id=2437753\", \"name\": \"RHBZ#2437753\", \"tags\": [\"issue-tracking\", \"x_refsource_REDHAT\"]}, {\"url\": \"https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-25640.json\", \"tags\": [\"x_sadp-csaf-vex\"]}], \"workarounds\": [{\"lang\": \"en\", \"value\": \"To reduce exposure, restrict network access to the Pydantic AI web interface. If the application is deployed on a remote server, configure firewall rules to limit access to trusted networks or localhost. This prevents untrusted clients from reaching the vulnerable component. After applying firewall changes, ensure the rules are persistent and the firewall service is reloaded.\"}], \"x_generator\": {\"engine\": \"sadp-cli 1.0.0\"}, \"descriptions\": [{\"lang\": \"en\", \"value\": \"A flaw was found in Pydantic AI. A remote attacker can exploit a path traversal vulnerability in the Pydantic AI web UI by crafting a malicious URL. This vulnerability arises from insufficient validation of the version query parameter, allowing the server to fetch and serve attacker-controlled HTML or JavaScript. If a victim clicks on such a link, malicious code can execute in their browser, potentially leading to the theft of sensitive client-side data, including chat history.\"}], \"problemTypes\": [{\"descriptions\": [{\"lang\": \"en\", \"type\": \"CWE\", \"cweId\": \"CWE-22\", \"description\": \"Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)\"}]}], \"providerMetadata\": {\"orgId\": \"0b0ca135-0b70-47e7-9f44-1890c2a1c46c\", \"shortName\": \"redhat-SADP\", \"dateUpdated\": \"2026-06-30T02:45:37.244Z\"}}, {\"title\": \"CISA ADP Vulnrichment\", \"metrics\": [{\"other\": {\"type\": \"ssvc\", \"content\": {\"id\": \"CVE-2026-25640\", \"role\": \"CISA Coordinator\", \"options\": [{\"Exploitation\": \"none\"}, {\"Automatable\": \"no\"}, {\"Technical Impact\": \"partial\"}], \"version\": \"2.0.3\", \"timestamp\": \"2026-02-06T20:16:17.763231Z\"}}}], \"providerMetadata\": {\"orgId\": \"134c704f-9b21-4f2e-91b3-4a467353bcc0\", \"shortName\": \"CISA-ADP\", \"dateUpdated\": \"2026-02-06T20:16:19.783Z\"}}], \"cna\": {\"title\": \"Pydantic AI affected by Stored XSS via Path Traversal in Web UI CDN URL\", \"source\": {\"advisory\": \"GHSA-wjp5-868j-wqv7\", \"discovery\": \"UNKNOWN\"}, \"metrics\": [{\"cvssV3_1\": {\"scope\": \"UNCHANGED\", \"version\": \"3.1\", \"baseScore\": 7.1, \"attackVector\": \"NETWORK\", \"baseSeverity\": \"HIGH\", \"vectorString\": \"CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N\", \"integrityImpact\": \"LOW\", \"userInteraction\": \"REQUIRED\", \"attackComplexity\": \"LOW\", \"availabilityImpact\": \"NONE\", \"privilegesRequired\": \"NONE\", \"confidentialityImpact\": \"HIGH\"}}], \"affected\": [{\"vendor\": \"pydantic\", \"product\": \"pydantic-ai\", \"versions\": [{\"status\": \"affected\", \"version\": \"\u003e= 1.34.0, \u003c 1.51.0\"}]}], \"references\": [{\"url\": \"https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7\", \"name\": \"https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7\", \"tags\": [\"x_refsource_CONFIRM\"]}, {\"url\": \"https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0\", \"name\": \"https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0\", \"tags\": [\"x_refsource_MISC\"]}], \"descriptions\": [{\"lang\": \"en\", \"value\": \"Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.34.0 to before 1.51.0,  a path traversal vulnerability in the Pydantic AI web UI allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. In affected versions, the CDN URL is constructed using a version query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data. This vulnerability only affects applications that use Agent.to_web to serve a chat interface and clai web to serve a chat interface from the CLI. These are typically run locally (on localhost), but may also be deployed on a remote server. This vulnerability is fixed in 1.51.0.\"}], \"problemTypes\": [{\"descriptions\": [{\"lang\": \"en\", \"type\": \"CWE\", \"cweId\": \"CWE-22\", \"description\": \"CWE-22: Improper Limitation of a Pathname to a Restricted Directory (\u0027Path Traversal\u0027)\"}]}, {\"descriptions\": [{\"lang\": \"en\", \"type\": \"CWE\", \"cweId\": \"CWE-79\", \"description\": \"CWE-79: Improper Neutralization of Input During Web Page Generation (\u0027Cross-site Scripting\u0027)\"}]}], \"providerMetadata\": {\"orgId\": \"a0819718-46f1-4df5-94e2-005712e83aaa\", \"shortName\": \"GitHub_M\", \"dateUpdated\": \"2026-02-06T20:01:53.892Z\"}}}",
      "cveMetadata": "{\"cveId\": \"CVE-2026-25640\", \"state\": \"PUBLISHED\", \"dateUpdated\": \"2026-06-30T02:45:37.244Z\", \"dateReserved\": \"2026-02-04T05:15:41.791Z\", \"assignerOrgId\": \"a0819718-46f1-4df5-94e2-005712e83aaa\", \"datePublished\": \"2026-02-06T20:01:53.892Z\", \"assignerShortName\": \"GitHub_M\"}",
      "dataType": "CVE_RECORD",
      "dataVersion": "5.2"
    }
  }
}

FKIE_CVE-2026-25640

Vulnerability from fkie_nvd - Published: 2026-02-06 20:16 - Updated: 2026-06-30 03:17

Severity

7.1 (High) - CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N
5.4 (Medium) - CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:L/A:N
7.1 (High) - CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N

Summary

References

URL	Tags
security-advisories@github.com	https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0	Product, Release Notes
security-advisories@github.com	https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7	Vendor Advisory
0b0ca135-0b70-47e7-9f44-1890c2a1c46c	https://access.redhat.com/security/cve/CVE-2026-25640
0b0ca135-0b70-47e7-9f44-1890c2a1c46c	https://bugzilla.redhat.com/show_bug.cgi?id=2437753
0b0ca135-0b70-47e7-9f44-1890c2a1c46c	https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-25640.json

Impacted products

	Vendor	Product	Version
	pydantic	pydantic_ai	*

JSON

To clipboard

{
  "affected": [
    {
      "affectedData": [
        {
          "product": "pydantic-ai",
          "vendor": "pydantic",
          "versions": [
            {
              "status": "affected",
              "version": "\u003e= 1.34.0, \u003c 1.51.0"
            }
          ]
        }
      ],
      "source": "security-advisories@github.com"
    },
    {
      "affectedData": [
        {
          "cpes": [
            "cpe:/a:redhat:enterprise_linux_ai:3"
          ],
          "defaultStatus": "affected",
          "product": "Red Hat Enterprise Linux AI (RHEL AI) 3",
          "vendor": "Red Hat"
        }
      ],
      "source": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c"
    }
  ],
  "configurations": [
    {
      "nodes": [
        {
          "cpeMatch": [
            {
              "criteria": "cpe:2.3:a:pydantic:pydantic_ai:*:*:*:*:*:python:*:*",
              "matchCriteriaId": "CD7C18E2-854E-44FD-A4AF-D4E580B67D98",
              "versionEndExcluding": "1.51.0",
              "versionStartIncluding": "1.34.0",
              "vulnerable": true
            }
          ],
          "negate": false,
          "operator": "OR"
        }
      ]
    }
  ],
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 1.34.0 to before 1.51.0,  a path traversal vulnerability in the Pydantic AI web UI allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. In affected versions, the CDN URL is constructed using a version query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data. This vulnerability only affects applications that use Agent.to_web to serve a chat interface and clai web to serve a chat interface from the CLI. These are typically run locally (on localhost), but may also be deployed on a remote server. This vulnerability is fixed in 1.51.0."
    },
    {
      "lang": "es",
      "value": "Pydantic AI es un framework de agente Python para construir aplicaciones y flujos de trabajo con IA Generativa. Desde la 1.34.0 hasta antes de la 1.51.0, una vulnerabilidad de salto de ruta en la interfaz de usuario web de Pydantic AI permite a un atacante servir JavaScript arbitrario en el contexto de la aplicaci\u00f3n al crear una URL maliciosa. En las versiones afectadas, la URL de la CDN se construye utilizando un par\u00e1metro de consulta de versi\u00f3n de la URL de la solicitud. Este par\u00e1metro no se valida, lo que permite secuencias de salto de ruta que hacen que el servidor obtenga y sirva HTML/JavaScript controlado por el atacante desde una fuente arbitraria en la misma CDN, en lugar del paquete leg\u00edtimo de la interfaz de usuario de chat. Si una v\u00edctima hace clic en el enlace o lo visita a trav\u00e9s de un iframe, el c\u00f3digo controlado por el atacante se ejecuta en su navegador, lo que permite el robo del historial de chat y otros datos del lado del cliente. Esta vulnerabilidad solo afecta a las aplicaciones que usan Agent.to_web para servir una interfaz de chat y clai web para servir una interfaz de chat desde la CLI. Estas suelen ejecutarse localmente (en localhost), pero tambi\u00e9n pueden implementarse en un servidor remoto. Esta vulnerabilidad se corrige en la 1.51.0."
    }
  ],
  "id": "CVE-2026-25640",
  "lastModified": "2026-06-30T03:17:43.313",
  "metrics": {
    "cvssMetricV31": [
      {
        "cvssData": {
          "attackComplexity": "LOW",
          "attackVector": "NETWORK",
          "availabilityImpact": "NONE",
          "baseScore": 7.1,
          "baseSeverity": "HIGH",
          "confidentialityImpact": "HIGH",
          "integrityImpact": "LOW",
          "privilegesRequired": "NONE",
          "scope": "UNCHANGED",
          "userInteraction": "REQUIRED",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N",
          "version": "3.1"
        },
        "exploitabilityScore": 2.8,
        "impactScore": 4.2,
        "source": "security-advisories@github.com",
        "type": "Secondary"
      },
      {
        "cvssData": {
          "attackComplexity": "LOW",
          "attackVector": "NETWORK",
          "availabilityImpact": "NONE",
          "baseScore": 5.4,
          "baseSeverity": "MEDIUM",
          "confidentialityImpact": "LOW",
          "integrityImpact": "LOW",
          "privilegesRequired": "NONE",
          "scope": "UNCHANGED",
          "userInteraction": "REQUIRED",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:L/A:N",
          "version": "3.1"
        },
        "exploitabilityScore": 2.8,
        "impactScore": 2.5,
        "source": "nvd@nist.gov",
        "type": "Primary"
      },
      {
        "cvssData": {
          "attackComplexity": "LOW",
          "attackVector": "NETWORK",
          "availabilityImpact": "NONE",
          "baseScore": 7.1,
          "baseSeverity": "HIGH",
          "confidentialityImpact": "HIGH",
          "integrityImpact": "LOW",
          "privilegesRequired": "NONE",
          "scope": "UNCHANGED",
          "userInteraction": "REQUIRED",
          "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N",
          "version": "3.1"
        },
        "exploitabilityScore": 2.8,
        "impactScore": 4.2,
        "source": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c",
        "type": "Secondary"
      }
    ],
    "ssvcV203": [
      {
        "source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
        "ssvcData": {
          "id": "CVE-2026-25640",
          "options": [
            {
              "exploitation": "none"
            },
            {
              "automatable": "no"
            },
            {
              "technicalImpact": "partial"
            }
          ],
          "role": "CISA Coordinator",
          "timestamp": "2026-02-06T20:16:17.763231Z",
          "version": "2.0.3"
        }
      }
    ]
  },
  "published": "2026-02-06T20:16:11.110",
  "references": [
    {
      "source": "security-advisories@github.com",
      "tags": [
        "Product",
        "Release Notes"
      ],
      "url": "https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0"
    },
    {
      "source": "security-advisories@github.com",
      "tags": [
        "Vendor Advisory"
      ],
      "url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7"
    },
    {
      "source": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c",
      "url": "https://access.redhat.com/security/cve/CVE-2026-25640"
    },
    {
      "source": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c",
      "url": "https://bugzilla.redhat.com/show_bug.cgi?id=2437753"
    },
    {
      "source": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c",
      "url": "https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-25640.json"
    }
  ],
  "sourceIdentifier": "security-advisories@github.com",
  "vulnStatus": "Modified",
  "weaknesses": [
    {
      "description": [
        {
          "lang": "en",
          "value": "CWE-22"
        },
        {
          "lang": "en",
          "value": "CWE-79"
        }
      ],
      "source": "security-advisories@github.com",
      "type": "Primary"
    },
    {
      "description": [
        {
          "lang": "en",
          "value": "CWE-22"
        }
      ],
      "source": "0b0ca135-0b70-47e7-9f44-1890c2a1c46c",
      "type": "Secondary"
    }
  ]
}

GHSA-WJP5-868J-WQV7

Vulnerability from github – Published: 2026-02-06 18:51 – Updated: 2026-02-06 21:43

Summary

Pydantic AI has Stored XSS via Path Traversal in Web UI CDN URL

Details

Summary

A Path Traversal vulnerability in the Pydantic AI web UI allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data.

This vulnerability only affects applications that use: - Agent.to_web to serve a chat interface - clai web to serve a chat interface from the CLI

These are typically run locally (on localhost), but may also be deployed on a remote server.

Description

The web UI serves its frontend HTML by fetching it from a CDN. In affected versions, the CDN URL is constructed using a version query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package.

Who Is Affected

Projects are affected if your application uses Agent.to_web or clai web to serve the Pydantic AI chat interface.

Attack Scenario

An attacker crafts a URL pointing to the victim's Pydantic AI web UI instance (either localhost with the known port, or a remote server endpoint) with a malicious version query parameter containing path traversal sequences.
The attacker gets the victim to visit this URL — directly via a link, through a redirect, or by embedding it in an iframe.
When the victim's browser loads the page, the server fetches and serves attacker-controlled HTML/JavaScript instead of the legitimate chat UI.
The attacker's JavaScript executes in the victim's browser in the context of the Pydantic AI web application, with access to:
Chat history stored in localStorage (all user messages and AI responses)
Session cookies that are not set as HttpOnly, if authentication middleware is configured

Remediation

Upgrade to Patched Version

Upgrade to the patched version or later. The fix removes the user-controllable version parameter entirely. The CDN URL is now hardcoded at startup and cannot be influenced by request parameters.

A new html_source parameter is available on Agent.to_web and create_web_app for applications that need to customize the UI source (e.g., for enterprise environments, offline usage, or custom UI builds). This parameter is only settable in application code, not via query parameters.

Severity

7.1 (High)


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

Show details on source website

JSON

To clipboard

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "pydantic-ai"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "1.34.0"
            },
            {
              "fixed": "1.51.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "pydantic-ai-slim"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "1.34.0"
            },
            {
              "fixed": "1.51.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-25640"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-22",
      "CWE-79"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-02-06T18:51:57Z",
    "nvd_published_at": "2026-02-06T20:16:11Z",
    "severity": "HIGH"
  },
  "details": "## Summary\n\nA Path Traversal vulnerability in the [Pydantic AI web UI](https://ai.pydantic.dev/web/) allows an attacker to serve arbitrary JavaScript in the context of the application by crafting a malicious URL. If a victim clicks the link or visits it via an iframe, attacker-controlled code executes in their browser, enabling theft of chat history and other client-side data.\n\n**This vulnerability only affects applications that use:**\n- **`Agent.to_web`** to serve a chat interface\n- **`clai web`** to serve a chat interface from the CLI\n\nThese are typically run locally (on `localhost`), but may also be deployed on a remote server.\n\n### Description\n\nThe web UI serves its frontend HTML by fetching it from a CDN. In affected versions, the CDN URL is constructed using a `version` query parameter from the request URL. This parameter is not validated, allowing path traversal sequences that cause the server to fetch and serve attacker-controlled HTML/JavaScript from an arbitrary source on the same CDN, instead of the legitimate chat UI package.\n\n### Who Is Affected\n\nProjects are affected if your application uses `Agent.to_web` or `clai web` to serve the Pydantic AI chat interface.\n\n### Attack Scenario\n\n1. An attacker crafts a URL pointing to the victim\u0027s Pydantic AI web UI instance (either `localhost` with the known port, or a remote server endpoint) with a malicious `version` query parameter containing path traversal sequences.\n\n2. The attacker gets the victim to visit this URL \u2014 directly via a link, through a redirect, or by embedding it in an iframe.\n\n3. When the victim\u0027s browser loads the page, the server fetches and serves attacker-controlled HTML/JavaScript instead of the legitimate chat UI.\n\n4. The attacker\u0027s JavaScript executes in the victim\u0027s browser in the context of the Pydantic AI web application, with access to:\n   - **Chat history** stored in `localStorage` (all user messages and AI responses)\n   - **Session cookies** that are not set as `HttpOnly`, if authentication middleware is configured\n\n## Remediation\n\n### Upgrade to Patched Version\n\n**Upgrade** to the patched version or later. The fix removes the user-controllable `version` parameter entirely. The CDN URL is now hardcoded at startup and cannot be influenced by request parameters.\n\nA new `html_source` parameter is available on `Agent.to_web` and `create_web_app` for applications that need to customize the UI source (e.g., for enterprise environments, offline usage, or custom UI builds). This parameter is only settable in application code, not via query parameters.",
  "id": "GHSA-wjp5-868j-wqv7",
  "modified": "2026-02-06T21:43:11Z",
  "published": "2026-02-06T18:51:57Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-wjp5-868j-wqv7"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-25640"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/pydantic/pydantic-ai"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pydantic/pydantic-ai/releases/tag/v1.51.0"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Pydantic AI has Stored XSS via Path Traversal in Web UI CDN URL"
}

Sightings

Author	Source	Type	Date	Other

Nomenclature

Seen: The vulnerability was mentioned, discussed, or observed by the user.
Confirmed: The vulnerability has been validated from an analyst's perspective.
Published Proof of Concept: A public proof of concept is available for this vulnerability.
Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
Not confirmed: The user expressed doubt about the validity of the vulnerability.
Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.

Detection rules are retrieved from Rulezet.

Action not permitted

CVE-2026-25640 (GCVE-0-2026-25640)

FKIE_CVE-2026-25640

GHSA-WJP5-868J-WQV7

Summary

Description

Who Is Affected

Attack Scenario

Remediation

Upgrade to Patched Version

Tags

Sightings

Nomenclature