HeadyZhang/agent-audit

Static security scanner for LLM agents — prompt injection, MCP config auditing, taint analysis. 49 rules mapped to OWASP Agentic Top 10 (2026). Works with LangChain, CrewAI, AutoGen.

52
/ 100
Established

Performs static taint analysis across tool boundaries to detect unsafe data flows into subprocess, eval, and SQL contexts, then correlates findings with framework-specific patterns (LangChain tool decorators, CrewAI agent definitions, AutoGen function signatures). Outputs results in SARIF, JSON, or baseline formats for CI integration, and includes a read-only MCP server inspector for auditing configuration and credential exposure without invoking tools.

104 stars. Available on PyPI.

Maintenance 13 / 25
Adoption 9 / 25
Maturity 18 / 25
Community 12 / 25

How are scores calculated?

Stars

104

Forks

11

Language

Python

License

MIT

Last pushed

Mar 11, 2026

Commits (30d)

0

Dependencies

6

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/agents/HeadyZhang/agent-audit"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.