If you're running AI agents in production and need structured accountability after each run, this gives you five tools to issue receipts, summarize what changed, check test evidence, flag review risks, and export audit logs. It's a remote MCP over streamable HTTP with bearer token auth, so you connect once and call it from your workflow orchestration. Clauxel hosts the endpoint and positions this for ops teams who want tamper-evident records without building their own ledger infrastructure. You'd reach for it when stakeholders ask "what did the agent actually do" and you need a machine-readable answer that survives rotation and handoffs.
claude mcp add --transport http clauxel-antigravity-run-ledger-mcp https://antigravityrunledger.clauxel.com/mcp