A lightweight, rule-based engine designed to automatically validate the quality and structure of AI agent outputs before they reach the user.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install self-review
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 self-review 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 Self-Review Skill?
The Self-Review Skill is an essential utility within the Openclaw Skills ecosystem, designed to ensure that AI-generated content meets high standards of professional communication. By leveraging a cost-effective, rule-based engine, it evaluates agent responses for clarity, conciseness, and actionability without incurring additional API costs. This skill serves as a critical quality gate, filtering out suboptimal content and ensuring that only refined, well-structured output is presented to the end user.
Self-Review Skill Use Cases
- Automating the quality assurance process for AI-driven customer support responses.
- Pruning verbose or redundant text from agent outputs to optimize token usage.
- Ensuring that technical documentation generated by agents follows specific structure and actionability standards.
- Integrating a sanity check step within complex multi-agent orchestration pipelines.
How Self-Review Skill Works
- The agent's generated text is piped into the Self-Review Skill via the command line or an automated pipeline step.
- The rule-based engine analyzes the input against predefined heuristics for clarity, brevity, and structural integrity.
- The skill evaluates the text and returns a status code: 0 for approval or 1 if the content requires further refinement.
- Based on the exit code, the workflow can either proceed to delivery or trigger a regeneration cycle.
Self-Review Skill Setup
To install and use this skill within your environment, follow these steps:
- Ensure you have Node.js installed.
- Clone the repository and navigate to the skill directory.
- Run the skill by piping output into the script:
echo "Generated content here" | node skills/self-review/index.js
To customize the evaluation logic, you can directly edit the threshold values within skills/self-review/index.js.
Self-Review Skill Data Schema & Taxonomy
The Self-Review Skill operates using a stream-based input model and provides binary feedback via standard exit codes. This makes it highly compatible with existing Openclaw Skills automation scripts.
| Component | Type | Description |
|---|---|---|
| Input | Stdin | The raw text string generated by the AI agent to be reviewed. |
| Approved | Exit Code 0 | Indicates the content passed all internal quality heuristics. |
| Rejected | Exit Code 1 | Indicates the content failed checks and requires improvement. |
| Config | index.js | Local configuration file for adjusting rule-based thresholds. |
Self-Review Skill Advanced Features
- Pipeline Integration: Seamlessly embed the skill as a standard step in your AGENTS.md workflow files.
- Zero API Overhead: Executes locally as a rule-based engine, providing an efficient and cost-free alternative to LLM-based reviewers.
- Customizable Heuristics: Developers can fine-tune the JavaScript logic to enforce specific brand voices or technical constraints.
- Ecosystem Compatibility: Designed to work alongside other Openclaw Skills, with a clear upgrade path to the self-review-llm package for deep semantic analysis.
SKILL.md
Loading
Related Openclaw Skills
Xiaohongshu Data Analytics
A comprehensive data collection and analysis toolkit for the Xiaohongshu platform designed to track notes, bloggers, and market trends.
Seisoai
A comprehensive media generation gateway designed for agents to dynamically discover and invoke state-of-the-art AI models for image, video, audio, and 3D assets.
Clawbridge
A lightweight chat integration to trigger Clawbridge discovery workflows and retrieve Vault links instantly.
RT DNA Finder: Football Player Similarity Search
Find statistically similar footballers and cheaper alternatives using advanced vector DNA similarity search.
Railway
A comprehensive tool for deploying and managing applications on the Railway cloud platform with zero configuration.
Baidu Search
An AI-powered search tool that enables agents to query the Baidu Search Engine for live web data, documentation, and research topics.
Featured*
