A utility skill to audit, classify, and clean up installed agent capabilities to prevent skill conflicts and clutter.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install workbuddy-skills-audit
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install workbuddy-skills-audit using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The Skills Audit skill is designed to maintain a clean and conflict-free environment for your AI agent capabilities. By scanning the local skills directory, it systematically identifies duplicate, legacy, platform-mismatched, or unapproved skills that may degrade performance or cause execution conflicts.
Optimizing your agent workflow with Openclaw Skills allows you to maintain high execution reliability. The tool compares your active capabilities against a golden reference list, ensuring your agent only utilizes verified and compatible workflows.
~/.workbuddy/skills/) to locate all active SKILL.md configurations.To start auditing your Openclaw Skills, use the bundled python script to generate a cleanup report.
# Run the default audit scan
python scripts/audit_skills.py
# Run with custom options
python scripts/audit_skills.py --dir ~/.workbuddy/skills/ --golden references/golden-list.md --output skills-audit-report.md
The skill generates a structured markdown report skills-audit-YYYY-MM.md containing a classification matrix. It references configuration and canonical lists to organize the data taxonomy.
| Status | Meaning | Action |
|---|---|---|
[OK] Approved |
Matches the canonical golden list | Keep without modification |
[!] Review |
Unknown skill not present in the golden list | Manually review validity |
[X] Platform: <name> |
Platform-specific skill (e.g., Clawdbot, Moltbot, OpenClaw) | Candidate for removal |
scripts/audit_skills.py: Execution script for directory analysis.references/golden-list.md: Canonical list of approved Openclaw Skills.--dir flag to audit alternative agent skillpaths.--golden.Loading
A high-fidelity research skill designed for multi-source validation, adversarial facts verification, and producing deeply-cited decision briefs.

A powerful local automation tool to capture WeChat public account articles, generate AI-driven summaries, and save them as structured Markdown notes in your Obsidian vault.

An AI-empowered dual-engine presentation generator built to design high-impact SAP enterprise-style, ESG, and digital finance slide decks in PPTX and Keynote formats.

A specialized intelligence compilation skill designed to generate decision-ready, structured 1-2 page executive research briefs.

An agent-to-agent cloud service that allows AI models to seamlessly deploy and run applications, manage custom domains, inspect logs, and monitor status.

An advanced preproduction planning skill designed to organize scripts, manage multi-character roles, and calibrate audiobook recordings before hitting the studio.








































