CVE-2026-133505 (GCVE-0-2026-133505)
Vulnerability from cvelistv5 – Published: 2026-11-20 07:34 – Updated: 2026-11-20 17:18
VLAI
EPSS
VEX
Title
Tool result data shaped like a file is rebuilt as a file download or provider file reference when message history is reloaded
Summary
When message history is serialized and loaded again, a tool result that contains a JSON object in the serialized shape of one of Pydantic AI's file types is rebuilt as that file. A third party who controls data a tool returns, such as a record that an MCP server or a function tool returns from an external system, can use this to make the application download a URL on its local network, or make the model provider read an s3:// object or uploaded file with the application's credentials. Applications are affected when tool results can contain third-party data, message history containing those results is serialized and loaded again (stored history, history a UI adapter holds on the server, or Temporal, DBOS or Prefect durable execution), and the model is one Pydantic AI downloads files for, one that honors force_download, or one whose provider resolves non-HTTP file URLs or uploaded files. History that a client submits through the Vercel AI or AG-UI adapter in the default mode is not affected. The issue is fixed in pydantic-ai and pydantic-ai-slim 1.107.10 and 2.66.0.
Severity
5.9 (Medium)
SSVC
Exploitation: none
Automatable: no
Technical Impact: partial
CISA Coordinator · CISA-ADP (v2.0.3)
Decision recorded 2026-11-20 17:18 UTC
References
2 references
| URL | Tags |
|---|---|
| https://github.com/pydantic/pydantic-ai/security/… | vendor-advisory |
| https://pydantic.dev/docs/ai/ui/overview/#trust-m… |
Impacted products
4 products
| Vendor | Product | Version | |
|---|---|---|---|
| pydantic | pydantic-ai |
Affected:
1.50.0 , ≤ 1.107.9
(python)
|
|
| pydantic | pydantic-ai |
Affected:
2.0.0b1 , ≤ 2.65.0
(python)
|
|
| pydantic | pydantic-ai-slim |
Affected:
1.50.0 , ≤ 1.107.9
(python)
|
|
| pydantic | pydantic-ai-slim |
Affected:
2.0.0b1 , ≤ 2.65.0
(python)
|
{
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-133505",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-11-20T17:18:02.957236Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-11-20T17:18:12.084Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai",
"product": "pydantic-ai",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "1.107.9",
"status": "affected",
"version": "1.50.0",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai",
"product": "pydantic-ai",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "2.65.0",
"status": "affected",
"version": "2.0.0b1",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai-slim",
"product": "pydantic-ai-slim",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "1.107.9",
"status": "affected",
"version": "1.50.0",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai-slim",
"product": "pydantic-ai-slim",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "2.65.0",
"status": "affected",
"version": "2.0.0b1",
"versionType": "python"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "When message history is serialized and loaded again, a tool result that contains a JSON object in the serialized shape of one of Pydantic AI\u0027s file types is rebuilt as that file. A third party who controls data a tool returns, such as a record that an MCP server or a function tool returns from an external system, can use this to make the application download a URL on its local network, or make the model provider read an s3:// object or uploaded file with the application\u0027s credentials. Applications are affected when tool results can contain third-party data, message history containing those results is serialized and loaded again (stored history, history a UI adapter holds on the server, or Temporal, DBOS or Prefect durable execution), and the model is one Pydantic AI downloads files for, one that honors force_download, or one whose provider resolves non-HTTP file URLs or uploaded files. History that a client submits through the Vercel AI or AG-UI adapter in the default mode is not affected. The issue is fixed in pydantic-ai and pydantic-ai-slim 1.107.10 and 2.66.0."
}
],
"metrics": [
{
"cvssV3_1": {
"baseScore": 5.9,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N",
"version": "3.1"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-843",
"description": "CWE-843: Access of Resource Using Incompatible Type",
"lang": "en",
"type": "CWE"
},
{
"cweId": "CWE-918",
"description": "CWE-918: Server-Side Request Forgery",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-20T07:34:58.052Z",
"orgId": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"shortName": "Pydantic"
},
"references": [
{
"tags": [
"vendor-advisory"
],
"url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-4cj4-hgwf-4hc4"
},
{
"url": "https://pydantic.dev/docs/ai/ui/overview/#trust-model-for-client-submitted-messages"
}
],
"source": {
"advisory": "GHSA-4cj4-hgwf-4hc4",
"discovery": "UNKNOWN"
},
"title": "Tool result data shaped like a file is rebuilt as a file download or provider file reference when message history is reloaded"
}
},
"cveMetadata": {
"assignerOrgId": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"assignerShortName": "Pydantic",
"cveId": "CVE-2026-133505",
"datePublished": "2026-11-20T07:34:58.052Z",
"dateReserved": "2026-11-19T23:17:25.405Z",
"dateUpdated": "2026-11-20T17:18:12.084Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2",
"vulnerability-lookup:meta": {
"epss": {
"cve": "CVE-2026-133505",
"date": "2026-11-22",
"epss": "0.00702",
"percentile": "0.5417"
},
"nvd": {
"cve": {
"affected": [
{
"affectedData": [
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai",
"product": "pydantic-ai",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "1.107.9",
"status": "affected",
"version": "1.50.0",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai",
"product": "pydantic-ai",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "2.65.0",
"status": "affected",
"version": "2.0.0b1",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai-slim",
"product": "pydantic-ai-slim",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "1.107.9",
"status": "affected",
"version": "1.50.0",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai-slim",
"product": "pydantic-ai-slim",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "2.65.0",
"status": "affected",
"version": "2.0.0b1",
"versionType": "python"
}
]
}
],
"source": "7cec5f2f-3032-4006-8740-31f605ffab5f"
}
],
"cveTags": [],
"descriptions": [
{
"lang": "en",
"value": "When message history is serialized and loaded again, a tool result that contains a JSON object in the serialized shape of one of Pydantic AI\u0027s file types is rebuilt as that file. A third party who controls data a tool returns, such as a record that an MCP server or a function tool returns from an external system, can use this to make the application download a URL on its local network, or make the model provider read an s3:// object or uploaded file with the application\u0027s credentials. Applications are affected when tool results can contain third-party data, message history containing those results is serialized and loaded again (stored history, history a UI adapter holds on the server, or Temporal, DBOS or Prefect durable execution), and the model is one Pydantic AI downloads files for, one that honors force_download, or one whose provider resolves non-HTTP file URLs or uploaded files. History that a client submits through the Vercel AI or AG-UI adapter in the default mode is not affected. The issue is fixed in pydantic-ai and pydantic-ai-slim 1.107.10 and 2.66.0."
}
],
"id": "CVE-2026-133505",
"lastModified": "2026-11-20T18:18:50.020",
"metrics": {
"cvssMetricV31": [
{
"cvssData": {
"attackComplexity": "HIGH",
"attackVector": "NETWORK",
"availabilityImpact": "NONE",
"baseScore": 5.9,
"baseSeverity": "MEDIUM",
"confidentialityImpact": "HIGH",
"integrityImpact": "NONE",
"privilegesRequired": "NONE",
"scope": "UNCHANGED",
"userInteraction": "NONE",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N",
"version": "3.1"
},
"exploitabilityScore": 2.2,
"impactScore": 3.6,
"source": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"type": "Secondary"
}
],
"ssvcV203": [
{
"source": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"ssvcData": {
"id": "CVE-2026-133505",
"options": [
{
"exploitation": "none"
},
{
"automatable": "no"
},
{
"technicalImpact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-11-20T17:18:02.957236Z",
"version": "2.0.3"
}
}
]
},
"published": "2026-11-20T08:18:33.637",
"references": [
{
"source": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-4cj4-hgwf-4hc4"
},
{
"source": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"url": "https://pydantic.dev/docs/ai/ui/overview/#trust-model-for-client-submitted-messages"
}
],
"sourceIdentifier": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"vulnStatus": "Received",
"weaknesses": [
{
"description": [
{
"lang": "en",
"value": "CWE-843"
},
{
"lang": "en",
"value": "CWE-918"
}
],
"source": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"type": "Secondary"
}
]
}
},
"redhat_vex": {
"aggregate_severity": "Moderate",
"current_release_date": "2026-11-20T11:28:50+00:00",
"cve": "CVE-2026-133505",
"id": "CVE-2026-133505",
"initial_release_date": "2026-11-20T07:34:58.052000+00:00",
"product_status:known_affected": "6",
"source": "Red Hat CSAF VEX",
"status": "final",
"title": "pydantic-ai: pydantic-ai-slim: pydantic-ai: Information disclosure via unsafe deserialization of tool results",
"url": "https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-133505.json",
"version": "3"
},
"vulnrichment": {
"containers": {
"adp": [
{
"metrics": [
{
"other": {
"content": {
"id": "CVE-2026-133505",
"options": [
{
"Exploitation": "none"
},
{
"Automatable": "no"
},
{
"Technical Impact": "partial"
}
],
"role": "CISA Coordinator",
"timestamp": "2026-11-20T17:18:02.957236Z",
"version": "2.0.3"
},
"type": "ssvc"
}
}
],
"providerMetadata": {
"dateUpdated": "2026-11-20T17:18:08.520Z",
"orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
"shortName": "CISA-ADP"
},
"title": "CISA ADP Vulnrichment"
}
],
"cna": {
"affected": [
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai",
"product": "pydantic-ai",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "1.107.9",
"status": "affected",
"version": "1.50.0",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai",
"product": "pydantic-ai",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "2.65.0",
"status": "affected",
"version": "2.0.0b1",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai-slim",
"product": "pydantic-ai-slim",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "1.107.9",
"status": "affected",
"version": "1.50.0",
"versionType": "python"
}
]
},
{
"collectionURL": "https://pypi.org",
"defaultStatus": "unaffected",
"packageName": "pydantic-ai-slim",
"product": "pydantic-ai-slim",
"repo": "https://github.com/pydantic/pydantic-ai",
"vendor": "pydantic",
"versions": [
{
"lessThanOrEqual": "2.65.0",
"status": "affected",
"version": "2.0.0b1",
"versionType": "python"
}
]
}
],
"descriptions": [
{
"lang": "en",
"value": "When message history is serialized and loaded again, a tool result that contains a JSON object in the serialized shape of one of Pydantic AI\u0027s file types is rebuilt as that file. A third party who controls data a tool returns, such as a record that an MCP server or a function tool returns from an external system, can use this to make the application download a URL on its local network, or make the model provider read an s3:// object or uploaded file with the application\u0027s credentials. Applications are affected when tool results can contain third-party data, message history containing those results is serialized and loaded again (stored history, history a UI adapter holds on the server, or Temporal, DBOS or Prefect durable execution), and the model is one Pydantic AI downloads files for, one that honors force_download, or one whose provider resolves non-HTTP file URLs or uploaded files. History that a client submits through the Vercel AI or AG-UI adapter in the default mode is not affected. The issue is fixed in pydantic-ai and pydantic-ai-slim 1.107.10 and 2.66.0."
}
],
"metrics": [
{
"cvssV3_1": {
"baseScore": 5.9,
"baseSeverity": "MEDIUM",
"vectorString": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:N/A:N",
"version": "3.1"
}
}
],
"problemTypes": [
{
"descriptions": [
{
"cweId": "CWE-843",
"description": "CWE-843: Access of Resource Using Incompatible Type",
"lang": "en",
"type": "CWE"
},
{
"cweId": "CWE-918",
"description": "CWE-918: Server-Side Request Forgery",
"lang": "en",
"type": "CWE"
}
]
}
],
"providerMetadata": {
"dateUpdated": "2026-11-20T07:34:58.052Z",
"orgId": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"shortName": "Pydantic"
},
"references": [
{
"tags": [
"vendor-advisory"
],
"url": "https://github.com/pydantic/pydantic-ai/security/advisories/GHSA-4cj4-hgwf-4hc4"
},
{
"url": "https://pydantic.dev/docs/ai/ui/overview/#trust-model-for-client-submitted-messages"
}
],
"source": {
"advisory": "GHSA-4cj4-hgwf-4hc4",
"discovery": "UNKNOWN"
},
"title": "Tool result data shaped like a file is rebuilt as a file download or provider file reference when message history is reloaded"
}
},
"cveMetadata": {
"assignerOrgId": "7cec5f2f-3032-4006-8740-31f605ffab5f",
"assignerShortName": "Pydantic",
"cveId": "CVE-2026-133505",
"datePublished": "2026-11-20T07:34:58.052Z",
"dateReserved": "2026-11-19T23:17:25.405Z",
"dateUpdated": "2026-11-20T17:18:12.084Z",
"state": "PUBLISHED"
},
"dataType": "CVE_RECORD",
"dataVersion": "5.2"
}
}
}
Loading…
Loading…
Experimental. This forecast is provided for visualization only and may change without notice. Do not use it for operational decisions.
Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.
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.
Loading…
Loading…
The MITRE ATT&CK techniques below are AI-generated suggestions, inferred from the description of the
vulnerability by the CIRCL/vulnerability-attack-technique-classification-roberta-base
model, served locally by ML-Gateway.
They have not been verified by an analyst and are provided for guidance only.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Loading…
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.
Loading…