Cryptographic audit trail for AI agents. Every event is hashed into an immutable SHA-256 chain. Detect tampering, prove what happened, comply with regulations — all from your MCP-compatible AI client.
Log events with hash chain — 15 event types, 5 severity levels
Retrieve full event details by ID
Search by agent, type, date range, or severity
Verify hash chain integrity between any two events
Get audit trail statistics and health metrics
Configure this MCP server in your AI client (Claude Desktop, Cursor, VS Code + Continue, etc.) and ask:
Describe natural language interactions your AI assistant can use with this server.
git clone https://github.com/Rumblingb/agent-audit-mcp.git
cd agent-audit-mcp
pip install -r requirements.txt
python3 server.py