A professional quality assurance tool designed to audit, score, and improve the code quality of AI agent skills based on official industry standards.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install agent-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 agent-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?
Skill Reviewer is a specialized utility for developers creating Openclaw Skills who want to ensure their agent instructions are robust, discoverable, and executable. It serves as a technical auditor that analyzes SKILL.md files for structural compliance, YAML frontmatter accuracy, and high-quality task-oriented instructions. By leveraging a structured scoring system, it helps maintain a high standard across the Openclaw Skills ecosystem.
The tool evaluates skills against the official Anthropic development guidelines, providing detailed feedback on description quality, example density, and error handling. Whether you are preparing a skill for private team use or public release, Skill Reviewer identifies critical flaws like broken code examples, misleading descriptions, and incorrect naming conventions before they impact your workflow.
Skill Reviewer Use Cases
- Validating a new skill's YAML frontmatter and folder structure before deployment.
- Conducting deep-dive audits for production-grade Openclaw Skills in professional environments.
- Improving the searchability of skills by optimizing description triggers and active verbs.
- Benchmarking example density to ensure AI agents have enough context to execute tasks.
- Quick-checking third-party Openclaw Skills before adding them to your local library.
How Skill Reviewer Works
- The user initiates a review by providing a skill path or the raw content of a SKILL.md file.
- The tool determines the audit mode (Regular or Strict) based on user intent and required quality standards.
- It performs a structural validation of the file system, checking for kebab-case naming and mandatory file presence.
- It parses the YAML frontmatter to verify required fields like name and description while ensuring no restricted keywords are used.
- It analyzes the instruction body for task-oriented organization and calculates the density of executable code examples.
- A comprehensive scorecard is generated, providing a percentage-based quality grade and a final recommendation (Publish, Revise, or Rework).
Skill Reviewer Setup
To use this tool within your agent environment, ensure you have the necessary reference guides available in your project structure:
# Verify the skill reviewer is in your skills directory
ls skills/skill-reviewer/
# Ensure reference guides are present for Strict Mode audits
ls skills/skill-reviewer/references/anthropic-skills-development-guide.md
# Run a structural check on a target skill
ls -la skills/your-target-skill/
Skill Reviewer Data Schema & Taxonomy
The skill organizes its evaluation logic into several key segments to maintain high-quality Openclaw Skills output:
| Category | Weight | Focus Area |
|---|---|---|
| Structural Check | 4pts | File naming, folder hierarchy, and kebab-case compliance. |
| Description Quality | 8pts | Active verbs, trigger conditions, and scope clarity. |
| Organization Audit | 6pts | Task-based grouping vs. abstract concept definitions. |
| Instruction Quality | 6pts | Step-by-step workflows and error handling. |
| Example Quality | 9pts | Code block density (ideal: 1 block per 5-30 lines). |
| Executability | 10pts | Imperative logic and verifiable output descriptions. |
| Tips & Expertise | 8pts | Non-obvious professional advice and pitfall warnings. |
Skill Reviewer Advanced Features
- Strict Mode: Mandatory alignment with full-length official development guides for production-level audits.
- Progressive Disclosure Check: Ensures documentation remains under 500 lines for optimal agent context management.
- Automatic Scorecard Generation: Outputs a formatted summary with a 100-point scale conversion.
- Metadata Validation: Checks for valid JSON formatting and accurate OS/Bin requirements in skill metadata.
- Sample Testing Workflow: Recommends a clean-shell test protocol for verifying 3-5 random commands from the skill.
SKILL.md
Loading
Related Openclaw Skills
Dream Interpreter
A local, zero-dependency reference tool providing comprehensive guides and best practices for dream analysis within your AI agent workflow.
KnowAir Forecast
KnowAir Forecast provides precise hourly and daily weather intelligence using the Caiyun Weather API for high-resolution global forecasting.
Exchange Rate
A script-based currency conversion tool providing real-time and historical exchange rates using European Central Bank data.
Create Discord Ticket Skill
This skill enables AI agents to automatically generate private Discord ticket channels to handle complex user issues and protect sensitive information.
AIPPT: Automated Presentation Generator
AIPPT is an automated skill that generates professional PowerPoint presentations from text descriptions or uploaded documents including docx and images.
Save Token
A specialized guiding skill that empowers AI agents to minimize token consumption by identifying and eliminating redundant conversation context.
Featured*
