MCP server for the nagi-ledger audit ledger and guardrail toolkit, exposing tools to record and annotate AI agent actions, subagent dispatches, and known dead ends, plus query session reports and statistics.
Local-first dashboard + MCP server that parses Claude Code and Codex JSONL files into a SQLite cost / token tracker. Per-MCP and per-tool breakdown, session drill-down, dedup by request_id; never talks to vendor APIs
Local MCP server that lets your AI coding agent query its own cross-tool project history - file/command freshness, past test failures, cost & token spend, cache status, and session handoff - over stdio, 100% local, no telemetry.
Enables MCP-compatible coding agents to debug applications using runtime log data, by providing tools to start a local log-ingestion server, track debugging sessions and hypotheses, and correlate logs to specific executions.