CVE-2026-48782

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. In versions 1.56.0 through 1.101.0, 2.0.0b1, and 2.0.0b2, the cloud-metadata blocklist could be bypassed by encoding the metadata IP in an IPv6 transition form that the previous fix, CVE-2026-46678...

CVE-2026-48782

CVE-2026-48782 affects Pydantic AI (versions 1.56.0–1.101.0, 2.0.0b1, 2.0.0b2) where the cloud-metadata blocklist can be bypassed by IPv6 transition forms that previous fixes did not decode. The IPv6 forms bypassing the blocklist can expose cloud IAM short-term credentials when an app uses force_...

PT-2026-50129

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. In versions 1.56.0 through 1.101.0, 2.0.0b1, and 2.0.0b2, the cloud-metadata blocklist could be bypassed by encoding the metadata IP in an IPv6 transition form that the previous fix, CVE-2026-46678...

antgent (>=0.3.0 <=0.3.2), ara-cli (>=0.1.14.13 <=0.1.14.14) +23 more potentially affected by CVE-2026-46678 via pydantic-ai (>=1.56.0 <=1.98.0)

pydantic-ai PYPI version =1.56.0, =0.3.0, =0.1.14.13, =1.5.0, =0.0.1, =1.0.0, =0.7.0rc1, =0.1.1, =0.1.0, =0.1.71, =0.0.1, =0.0.1, =0.0.8 - mseep-zoekt-mcp =0.1.0 - murmur-runtime =0.1.0 and more Source cves: CVE-2026-46678 Source advisory: OSV:GHSA-CQP8-FCVH-X7R3...

GHSA-CQP8-FCVH-X7R3 Pydantic AI: SSRF cloud-metadata blocklist bypass via IPv4-mapped IPv6 (Incomplete fix of CVE-2026-25580)

Summary When an application using Pydantic AI opts a URL into forcedownload='allow-local' which disables the default block on private/internal IPs, the cloud-metadata blocklist could be bypassed by encoding the metadata IP in an IPv6 transition form IPv4-mapped IPv6, 6to4, or NAT64. Dual-stack an...

antgent (>=0.3.0 <=0.3.2), ara-cli (>=0.1.14.13 <=0.1.14.14) +44 more potentially affected by CVE-2026-46678 via pydantic-ai-slim (>=1.56.0 <=1.98.0)

pydantic-ai-slim PYPI version =1.56.0, =0.3.0, =0.1.14.13, =1.5.0, =0.1.0a1, =0.0.400, =0.0.1, =1.0.0, =1.0.3, =0.0.498, =0.1.1, =0.7.0rc1, =0.1.1, =0.1.0, =0.3.1 and more Source cves: CVE-2026-46678 Source advisory: OSV:GHSA-CQP8-FCVH-X7R3...

antgent (>=0.3.0 <=0.3.2), ara-cli (>=0.1.14.13 <=0.1.14.14) +44 more potentially affected by CVE-2026-25580 +1 more via pydantic-ai-slim (>=1.56.0 <=1.98.0)

pydantic-ai-slim PYPI version =1.56.0, =0.3.0, =0.1.14.13, =1.5.0, =0.1.0a1, =0.0.400, =0.0.1, =1.0.0, =1.0.3, =0.0.498, =0.1.1, =0.7.0rc1, =0.1.1, =0.1.0, =0.3.1 and more Source cves: CVE-2026-25580, CVE-2026-46678 Source advisory: SNYK:PYTHON-PYDANTICAISLIM-16796278...

Server-side Request Forgery (SSRF)

Overview pydantic-ai-slim is an Agent Framework / shim to use Pydantic with LLMs, slim package Affected versions of this package are vulnerable to Server-side Request Forgery SSRF via incomplete blocklist in isprivateip function when forcedownload='allow-local' is enabled. An attacker can access...

sec-recon-agent

sec-recon-agent Type-safe security triage built on Pydantic A...

Server-Side Request Forgery (SSRF)

Pydantic AI is vulnerable to Server-Side Request Forgery SSRF. The vulnerability is due to improper validation of URLs in the download functionality when processing untrusted message history, which allows an attacker to supply malicious URLs that force the server to make unauthorized requests to...

CVE-2026-25904

The Pydantic-AI MCP Run Python tool configures the Deno sandbox with an overly permissive configuration that allows the underlying Python code to access the localhost interface of the host to perform SSRF attacks. Note - the "mcp-run-python" project is archived and unlikely to receive a fix...

CVE-2026-25904

The Pydantic-AI MCP Run Python tool configures the Deno sandbox with an overly permissive configuration that allows the underlying Python code to access the localhost interface of the host to perform SSRF attacks. Note - the "mcp-run-python" project is archived and unlikely to receive a fix...

CVE-2026-25904

The Pydantic-AI MCP Run Python tool configures the Deno sandbox with an overly permissive configuration that allows the underlying Python code to access the localhost interface of the host to perform SSRF attacks. Note - the "mcp-run-python" project is archived and unlikely to receive a fix...

Cross-site Scripting

Pydantic AI is vulnerable to Cross-Site Scripting XSS. The vulnerability is due to insufficient validation of a user-controlled version parameter when constructing CDN URLs, allowing path traversal that causes the server to fetch and serve attacker-controlled JavaScript from the CDN, leading to...

CVE-2026-25640

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 HT...

CVE-2026-25580

A flaw was found in Pydantic AI. This Server-Side Request Forgery SSRF vulnerability allows a remote attacker to include malicious URLs within untrusted message history. When processed by the application, these URLs can force the server to make unauthorized HTTP requests to internal network...

CVE-2026-25580 Pydantic AI Affected by Server-Side Request Forgery (SSRF) in URL Download Handling

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.0.26 to before 1.56.0, aServer-Side Request Forgery SSRF vulnerability exists in Pydantic AI's URL download functionality. When applications accept message history from untrusted sources,...

CVE-2026-25580

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.0.26 to before 1.56.0, aServer-Side Request Forgery SSRF vulnerability exists in Pydantic AI's URL download functionality. When applications accept message history from untrusted sources,...

CVE-2026-25580 Pydantic AI Affected by Server-Side Request Forgery (SSRF) in URL Download Handling

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.0.26 to before 1.56.0, aServer-Side Request Forgery SSRF vulnerability exists in Pydantic AI's URL download functionality. When applications accept message history from untrusted sources,...

CVE-2026-25580 Pydantic AI Affected by Server-Side Request Forgery (SSRF) in URL Download Handling

Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. From 0.0.26 to before 1.56.0, aServer-Side Request Forgery SSRF vulnerability exists in Pydantic AI's URL download functionality. When applications accept message history from untrusted sources,...