Skip to content

Some text is shown in its original language while the English translation is being prepared.

EventMonitoring

Pydantic AI fixes web_fetch denial-of-service vulnerability

1 reports1 reporting sourcesLatest report

Start here

AI overview

Pydantic AI v2.52.0 fixes a medium-severity vulnerability in the local web_fetch tool (GHSA-v36g-jcw9-x7cw): processing attacker-controlled deeply nested HTML consumes excessive CPU and memory. Provider-native fetching is unaffected. The vulnerability was fixed in 2.52.0 (v2) and 1.107.7 (v1). This version also adds support for Claude Sonnet 5.5 and OpenAI gpt-6.1-sol, and adjusts the behavior of AnthropicModel under Claude-related configurations.

Generated by AI from reports · Updated 3 days ago

Latest update
Pydantic AI v2.52.0 修复 web_fetch 拒绝服务漏洞并加入 Claude Sonnet 5.5 支持

Progress timeline

Oct 10
  1. Pydantic AI
    Pydantic AI v2.52.0 修复 web_fetch 拒绝服务漏洞并加入 Claude Sonnet 5.5 支持

    Pydantic AI v2.52.0 修复了本地 web_fetch 工具的一处中危漏洞（GHSA-v36g-jcw9-x7cw）：处理攻击者控制的深层嵌套 HTML 会消耗过多 CPU 和内存，provider 原生抓取不受影响，已在 2.52.0（v2）和 1.107.7（v1）修复。该版本还新增 Claude Sonnet 5.5 与 OpenAI gpt-6.1-sol 支持，并将 AnthropicModel 在 Claude Sonnet 4.5 及之后的默认 max_tokens 从 4096 提升到 16384。

Related discussion

No discussion links have been collected for this event yet.

Attention trend

New sources per day

3 days total1 ·Maximum1（Oct 10）·Today0

  • Oct 10New1 sources (1 outlets,0 accounts)
  • Oct 11New0 sources (0 outlets,0 accounts)
  • Oct 12New0 sources (0 outlets,0 accounts)

Hourly attention

No comparable trend data