clawmetry and openclaw-inspector
One tool provides real-time observability into OpenClaw AI agent thought processes, while the other offers post-hoc review and flagging of past conversations, making them complementary for comprehensive agent monitoring.
About clawmetry
vivekchand/clawmetry
See your agent think. Real-time observability dashboard for OpenClaw AI agents.
Provides live animated message flows across 18+ channels (Telegram, Discord, Slack, WhatsApp, Signal, Matrix, etc.), token/cost tracking by model and session, and workspace memory browsing—all auto-detected from `openclaw.json` with zero configuration. Built on Flask, it runs locally on port 8900 and streams real-time logs, cron job status, and agent event traces through a unified dashboard UI.
About openclaw-inspector
Lukavyi/openclaw-inspector
Review every conversation your OpenClaw agent has ever had — including deleted sessions — and flag dangerous actions.
Provides real-time monitoring and forensic analysis of OpenClaw agent activity through a local web interface, parsing JSONL session histories to detect dangerous patterns like destructive filesystem commands, credential exposure, and surveillance actions with customizable severity rules. Includes multi-axis filtering, read-progress tracking, and live SSE updates across active, orphan, and deleted sessions—all entirely localhost-bound with zero cloud dependencies or external telemetry.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work