pydantic-ai-slim is vulnerable to Denial of Service (DoS)
75
High Risk
ConcurrencyLimitedModel and limit_model_concurrency acquire a slot on a shared anyio.CapacityLimiter on the task that starts a model stream, then release that slot from a different internal cleanup task when the stream ends early or finishes through the default stream_text() debouncing. The limiter rejects a release from a non-owning task, so the slot stays held after the request completes. Repeatedly starting and abandoning streaming requests against a shared limiter exhausts its slots and blocks every later request that shares it. The fix keeps slot acquisition and release on the same task.
You are affected if you are using a version that falls within the vulnerable range and you share a ConcurrencyLimitedModel or limit_model_concurrency limiter across streaming model requests.
pydantic-ai-slim is vulnerable to Denial of Service (DoS) in versions 2.10.0 - 2.52.0.
Upgrade the pydantic-ai-slim and/or the pydantic-ai library to the patch version.
Connect your repositories to instantly see whether vulnerable or malicious packages exist in your codebase.
Free. No credit card required.

I consent to receiving marketing communications based on Aikido’s Privacy Policy.
SOC 2Compliant
ISO 27001Compliant
ISO 42001Compliant© All Intel data is openly available and commercially licensed.