ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified developer tools Safety 5/5

evidence-based-investigation

Investigate problems using evidence-based analysis with session logs, error logs, and 5 Whys methodology. Use when debugging OpenClaw issues, analyzing system failures, tracing message delivery problems, or conducting root cause analysis that requires hard evidence from logs, files, or system state rather than assumptions.

Why use this skill?

Master your debugging process with the Evidence-Based Investigation skill. Use logs, system state, and 5 Whys to fix OpenClaw errors.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/chunhualiao/evidence-based-investigation
Or

What This Skill Does

The evidence-based-investigation skill provides a rigorous, systematic framework for diagnosing issues within the OpenClaw environment. Rather than relying on intuition or trial-and-error, this skill enforces a strict protocol requiring users to identify observable problems, extract raw data from session and system logs, and apply the '5 Whys' methodology to determine the actual root cause. It forces a critical review of initial decision-making processes before diving into code or configuration, ensuring that time is not wasted troubleshooting an issue that was actually a user-level selection error.

Installation

To install this skill, run the following command in your OpenClaw terminal: clawhub install openclaw/skills/skills/chunhualiao/evidence-based-investigation

Use Cases

  • Debugging message delivery failures between agents and gateways.
  • Performing root cause analysis on system crashes using error logs.
  • Investigating why an agent chose a specific tool or workflow path.
  • Tracing data inconsistencies back to specific session events.
  • Verifying if a failed task was due to an environmental issue or an incorrect tool invocation.

Example Prompts

  1. "I am seeing constant errors when the agent tries to parse JSON files. Use evidence-based-investigation to find out why the parser is failing in session 8829."
  2. "My agent keeps trying to use a browser fetcher on a binary file. Perform an investigation to see if this was a decision point error or a system failure."
  3. "Investigate the message delivery failure from 2026-02-14 at 23:18. Use gateway logs and session history to determine the root cause using the 5 Whys."

Tips & Limitations

  • Tip: Always start your investigation by checking the Decision Point Analysis section of the documentation. Many 'bugs' are actually configuration or tool-selection mistakes.
  • Tip: Utilize the provided jq and grep snippets to filter massive log files before analyzing them to prevent information overload.
  • Limitation: This skill requires access to underlying system files. If you do not have appropriate permissions to read logs, the investigation will be limited to session-level data only. Always ensure you are working within the correct directory path for your specific OpenClaw agent deployment.

Metadata

Stars1100
Views0
Updated2026-02-17
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-chunhualiao-evidence-based-investigation": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#debugging#root-cause-analysis#logging#troubleshooting
Safety Score: 5/5

Flags: file-read