A comprehensive auditing framework and scoring system for evaluating the quality, structure, and actionability of AI agent skill files.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-reviewer
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install skill-reviewer using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Skill Reviewer?
The Skill Reviewer is a specialized utility designed to maintain high standards for AI agent capabilities by auditing SKILL.md files. It provides a structured framework to evaluate every aspect of a skill, from valid YAML frontmatter and searchable descriptions to code block density and cross-platform accuracy. By using this tool within the Openclaw Skills ecosystem, developers can ensure their agents receive clear, actionable, and error-free instructions.
This skill acts as a quality gate for both developers publishing new skills and users evaluating third-party integrations. It transforms raw markdown documentation into a high-performance technical asset by identifying common defects like placeholder values, broken code examples, and abstract organization that hinders agent effectiveness.
Skill Reviewer Use Cases
- Pre-publication auditing of new agent skills to ensure registry readiness.
- Evaluating third-party skills downloaded from the registry for safety and quality.
- Benchmarking and comparing multiple skills within the same category for performance.
- Identifying and fixing technical debt in existing Openclaw Skills.
- Scoring the actionability of agent instructions to reduce runtime errors.
How Skill Reviewer Works
- Perform a structural check to validate YAML frontmatter, required fields like name and metadata, and basic document hierarchy.
- Audit the description and metadata for SEO optimization, trigger phrase inclusion, and accurate tool requirement listing.
- Analyze content quality by calculating code-to-text density and verifying the syntax of provided examples.
- Assess actionability by checking if instructions are imperative and logically ordered for an AI agent to execute.
- Evaluate the tips section for non-obvious expertise and specific gotcha coverage.
- Generate a final scorecard with a weighted rating to determine if the skill is ready for publishing or requires rework.
Skill Reviewer Setup
To begin auditing with this tool, ensure you have the necessary CLI utilities installed. You can install skills for review using the following command:
npx molthub@latest install skill-name
To manually inspect the skill structure and check for common defects like placeholders or missing language tags, use standard Unix utilities:
# Check for placeholders like 'foo' or 'TODO'
grep -n -i 'todo\\|fixme\\|xxx\\|foo\\|bar\\|baz' SKILL.md
# Validate language tags in code blocks
grep -n '^```$' SKILL.md
Skill Reviewer Data Schema & Taxonomy
The Skill Reviewer uses a structured scorecard to quantify quality. Data is organized across the following dimensions:
| Category | Max Score | Key Focus Area |
|---|---|---|
| Structure | 11 | YAML validity and mandatory sections |
| Description | 8 | Active verbs, trigger phrases, and searchability |
| Content | 9 | Example density (1 block per 8-15 lines) and syntax |
| Actionability | 10 | Imperative steps and error handling |
| Tips | 8 | Non-obvious expertise and specific gotchas |
Scores are aggregated to provide a final verdict: Excellent (45+), Good (35-44), Fair (25-34), or Poor (<25).
Skill Reviewer Advanced Features
- Automated defect detection for critical blockers like broken YAML or missing tool requirements.
- Comparative review framework for selecting the best Openclaw Skills among similar options.
- Checklist-driven Quick Review template for rapid evaluation of agent skills.
- Cross-platform accuracy auditing to ensure compatibility across macOS, Linux, and Windows.
- Metric-based quality signaling using code-to-text density ratios.
SKILL.md
Loading
Related Openclaw Skills
Shell Scripting
A professional toolkit for building reliable, portable, and production-ready Bash and POSIX shell scripts.
Security Audit
A comprehensive security scanner for auditing codebases, detecting leaked secrets, and identifying infrastructure vulnerabilities.
Regex Patterns
A comprehensive regular expression cookbook for input validation, data extraction, and code refactoring across multiple programming languages.
Performance Profiler
A comprehensive performance diagnostics toolkit for profiling CPU usage, detecting memory leaks, and benchmarking application code across multiple languages.
Skill Search Optimizer
A technical guide and toolset for optimizing agent skill discoverability through semantic search alignment and description engineering.
Skill Writer
A comprehensive framework for writing and publishing standardized Markdown-based skills that AI coding agents use to execute tasks and follow technical instructions.
Featured*
