Mingjing— AI Agent Health Center
ReviewAudited by ClawScan on May 13, 2026.
Overview
The skill’s documented monitoring, local storage, and dashboard behavior fit its stated health-monitoring purpose, but users should verify the external package/plugin and understand what agent telemetry is stored locally.
Before installing, verify the `mingjing` package and `mingjing-probe` plugin source, understand what telemetry is collected and stored locally, and make sure you can stop the daemon and dashboard when you are done.
Publisher note
Network: none — zero outbound traffic, data stays local LLM: none — pure rule engine, zero model calls Permissions: filesystem:write (to ~/.ming/ for hot logs and SQLite) Sandbox: safe — passive observer, no code execution
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
You would be trusting code from an external package source that was not included in this review.
The skill instructs users to install an external Python package, but the submitted artifact set contains only SKILL.md and no reviewed package source or install spec.
pip install mingjing
Verify the PyPI package and GitHub repository before installing, and prefer pinned versions or trusted release artifacts.
The plugin may receive access to agent runtime events, so its behavior matters even though the request is aligned with the monitoring purpose.
The skill asks users to install and enable an OpenClaw probe plugin, but the plugin implementation is not included in the provided artifacts.
openclaw plugins install mingjing-probe
Review the plugin source or package provenance before enabling it in OpenClaw.
Local monitoring data could include sensitive prompts, outputs, tool results, or agent memory details, even if it is not sent over the network.
The skill discloses local storage of monitoring data. Its feature tables also describe live event streams and diagnostics covering model output, tool output, and memory retrieval/storage.
Offline-first | No outbound traffic, data stored locally
Check where logs are stored, how long they are retained, and whether sensitive content is redacted before using it on private workflows.
The monitoring service may keep running after setup and expose diagnostics on a local web port.
The documented workflow starts a background daemon and a local web dashboard, which is expected for real-time monitoring but creates persistent local processes.
python3 -m src.ming start --daemon python3 -m src.ming web start --port 18088
Only start it when needed, confirm it binds locally, and learn how to stop the daemon and web panel.