Live › vulnerability › CVE-2026-114386
Go to dashboardPydantic AI Harness before 0.59.0 does not enforce the tool-call limit on tools invoked from nested calls inside its CodeMode sandbox. A user or prompt-injected content can make the sandboxed program call tools beyond the operator's configured limit, including costly or side-effecting ones.
Summary written from the sources below. Check the originals before acting.
First seen by InfraStack 2026-10-23 20:59 UTC, 1h before NVD. All times on this page are UTC.
| Product | Ecosystem | Affected | Fixed in |
|---|---|---|---|
| pydantic-ai-harness | pypi | < 0.59.0 | 0.59.0 |
Check whether the project pins pydantic-ai-harness below 0.59.0 (e.g. pip show pydantic-ai-harness or grep it in requirements/lock files).
Upgrade pydantic-ai-harness to 0.59.0 or later.
Each row is the first time that source carried this signal. The analyst’s rows are its own findings.
Pydantic AI Harness before 0.59.0 fails to enforce UsageLimits.tool_calls_limit on nested tool calls dispatched from run_code in the CodeMode toolset. Users or prompt-injected content can make the sandboxed program invoke tools via run_nested_call() without check_before_tool_call, exceeding operator limits on costly or side-effecting tools.
This page as .md · .json · in your agent: get_signal("CVE-2026-114386")
GitHub Advisory Database, CC-BY 4.0 · NVD data is provided by NIST and is public domain. Posts quoted belong to their authors.