Agent Audit Logger Mcp
Register this MCP when your Claude Code or Cursor agents call each other or external tools and you need a tamper-evident audit trail for compliance, incident review, or trust in multi-agent workflows.
Overview
Agent Audit Logger MCP is an Operate-phase MCP server that writes hash-chained, HMAC-signed audit logs for agent-to-agent tool calls and handshakes.
What is this MCP server?
- Hash-chained HMAC-signed audit records for agent-to-agent (A2A) tool calls and related agent handshakes
- Stdio MCP server (Python, PyPI package agent-audit-logger-mcp v1.0.4)
- Designed for tamper-evident trails rather than ad hoc console logging
- Fits regulated or multi-agent setups where non-repudiation matters
- Complements security review workflows without replacing human code review
- Published version 1.0.4 on PyPI (identifier agent-audit-logger-mcp)
- Transport: stdio
- Runtime hint: Python (registryType pypi)
What problem does it solve?
When agents call tools and each other in production, plain logs are easy to tamper with and painful to defend in an audit or outage.
Who is it for?
Solo builders running multi-step agent workflows who need a dedicated A2A audit MCP on stdio with a published Python package.
Skip if: Teams that only need local dev tracing with no integrity requirements, or builders without any multi-agent or tool-call surface to record.
What do I get? / Deliverables
After install, your agent can append cryptographically linked audit events you can replay and verify when debugging or proving compliance.
- Stdio MCP endpoint for append/query of hash-chained audit events
- Verifiable records covering tool calls and agent handshakes in A2A flows
- Foundation for retention policies and external SIEM export you configure
Recommended MCP Servers
Journey fit
Audit logging becomes critical once agents run in production and exchange tool calls; the Operate phase is where builders prove what happened and meet governance expectations. Monitoring is the canonical shelf for hash-chained, signed logs that support ongoing observability and post-incident reconstruction of agent-to-agent (A2A) activity.
How it compares
Tamper-evident audit MCP integration, not a general application logger or a static security checklist skill.
Common Questions / FAQ
Who is agent-audit-logger-mcp for?
Indie and solo developers operating Claude Code, Cursor, or similar agents that delegate work agent-to-agent and need signed, chain-linked audit evidence.
When should I use agent-audit-logger-mcp?
Use it when agents are live, tools are shared across agents, or you must show a defensible record of tool calls and handshakes beyond unstructured logs.
How do I add agent-audit-logger-mcp to my agent?
Install the PyPI package agent-audit-logger-mcp (v1.0.4), register the stdio MCP server in Claude Code or Cursor, and configure signing keys and storage per the GitHub repository README.