Automates the evaluation of pull requests for AI coding agent skills to guarantee architectural validity, design rigor, and seamless marketplace compatibility.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install review-skill
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 review-skill 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 Review Skill?
The review-skill provides an automated, programmatic framework to evaluate pull requests adding or modifying Openclaw Skills. By checking structural specifications, metadata design quality, and ecosystem-wide marketplace compliance, it prevents regression and guarantees that newly submitted skills parse seamlessly.
Operating with a strict verification protocol, this skill enforces a tiered severity calibration system. It removes human bias by categorizing violations with explicit confidence levels, enabling continuous integration pipelines to mechanically block problematic merges before they impact production environments.
Review Skill Use Cases
- Automated CI/CD Quality Gates: Run automated assessments during git pull request pipelines to verify submitted skills.
- Ecosystem Manifest Auditing: Scan existing repository structures for outdated paths, formatting anomalies, or naming collisions.
- Pre-publication Validation: Execute comprehensive local evaluations prior to pushing new open-source contributions to the shared registry.
How Review Skill Works
- Parse Arguments: Extract the targets and optional configurations from the provided arguments, falling back to a default output path and the main branch if unspecified.
- Identify Changed Skills: Execute git diff comparisons against the base branch to filter out unchanged files and target specific skill directories.
- Read Files End-to-End: Ingest the complete SKILL.md documentation, underlying script resources, and corresponding marketplace manifests.
- Run Validation Multi-Pass: Evaluate the files through sequentially layered structural, design, and marketplace verification protocols.
- Compile Verified Output: Aggregate all confirmed structural findings into a formatted markdown ledger complete with explicit line numbers, impact explanations, and a definitive readiness verdict.
Review Skill Setup
Integrate the review-skill directly into your environment using the command line:
# Run the review skill against your modified workspace
openclaw run review-skill --arguments ".review-output.md" --base "main"
Review Skill Data Schema & Taxonomy
The skill generates an output markdown ledger cataloging all system validations categorized by rigorous operational thresholds:
| Severity Level | Focus Areas | Merge Impact |
|---|---|---|
| Critical | Missing or malformed YAML frontmatter, name collisions, or file length exceeding 500 lines | Blocks Approval |
| Major | Incorrect casing formats, overly vague description components, or nested references | Blocks Approval |
| Minor | Hardcoded temporal dates, inconsistent taxonomy, or missing local cross-references | Informational Only |
| Informational | Potential trigger phrase overlaps or non-standard interactive scripting hints | Informational Only |
Review Skill Advanced Features
- Verification Protocol Guardrails: Cross-references findings with standard verification policies to completely strip away subjective style preferences and maintain deterministic evaluation.
- Subsequent Re-Review Safeguards: Implements strict iterative pass boundaries designed to focus only on checking prior failures while preventing discovery creep.
- Confidence-Scored Metadata: Attaches high or medium diagnostic confidence metrics to every tracked issue based on whether the parameter is mechanically or subjectively verified.
SKILL.md
Loading
Related Openclaw Skills
Review Skill Improver
An automated analytical tool designed to optimize review quality by diagnosing rule discrepancies and proposing targeted refinements based on feedback logs.
Remix v2 Routing
A specialized AI skill to master Remix v2 routing patterns, including flat-routes conventions, nested layouts, and resource routes.
Remix v2 Performance & SSR Code Review
An automated code review agent skill designed to audit Remix v2 applications for hydration mismatches, server/client code leaks, suboptimal caching headers, and asset delivery bottlenecks.
Review Structure
An advanced codebase maintainability and structural review skill that helps AI agents identify and execute code-judo restructurings while guarding against architectural drift.
Uiverse Design
A curated library of 165+ open-source CSS and Tailwind CSS UI components designed to help developers and AI agents browse, search, and integrate beautiful UI components from Uiverse Galaxy.
Fresh Graduate Resume Writer & Optimization Assistant
An AI-powered assistant designed to help fresh graduates build, polish, and optimize job-winning resumes tailored to modern HR requirements.
Featured*
