CVE-2026-65975
CVE-2026-65975: 1 tracked advisory record across Red Hat. Compare vendor sources and published fix guidance.
Compare the source-linked records below. Ratings and product/version details belong to each advisory; they are not a single CVE-wide score or proof that every listed product is affected. How VulniPulse collects and checks evidence.
Vendor advisory comparison
Red Hat
1 advisory- Advisory severityMedium6.5
Medium [CVE-2026-65975] Remote clients can execute server tools with forged arguments due to improper message sanitization.
CVE-2026-65975Source published Source updated
Pydantic AI is a Python agent framework for building applications and workflows with Generative AI. In versions 1.88.0 up to but not including 1.107.1 and 2.0.0b1 up to but not including 2.5.0, the UI adapters (AG-UI via Agent.to_ag_ui()/AGUIAdapter, and Vercel AI via VercelAIAdapter) use sanitize_messages to strip unresolved ("dangling") client-submitted tool calls from untrusted message history before it reaches the agent, a defense-in-depth default that prevents the agent from executing tool calls the model never emitted. However, the strip anchored to a message index computed before sanitization ran, so when a trailing client message sanitized to empty and was dropped (for example a client system message under the default manage_system_prompt='server'), a preceding assistant response…
- Affected products in this advisory
- Lightspeed Core
- Source-reported affected versions
- < 1.88.0
- < 1.107.1
- < 2.0.0b1
- < 2.5.0
- Source-reported fixed versions
- 1.107.1
- 2.5.0
- Mitigation guidance
- Configure Pydantic AI tools to require explicit approval for execution. This prevents remote clients from automatically executing server tools with forged arguments, as approval-gated tools are not affected by this vulnerability. Additionally, restrict network access to the Pydantic AI application to trusted clients only, if feasible in your environment.
Android app · Google Play
Monitor future Red Hat CVEs from your phone.
Choose a whole vendor or a precise platform, then receive matching security advisories by phone notification, email, or both. Coverage follows 34 official vendor sources and 160+ reviewed platform categories.