A meta-skill designed to evaluate and optimize other skills by identifying gaps in logic, codebase fit, and output quality.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install improve-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 improve-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 Improve Skill: The Meta-Skill?
The Improve Skill serves as a continuous improvement engine for your development environment. It functions as a meta-skill that audits other Openclaw Skills against the specific requirements of your current project. By analyzing skill definitions, supporting files, and invocation history, it ensures that your automation tools remain effective even as your codebase evolves.
This tool is essential for developers who want to tighten the alignment between their AI agents and their project architecture. It bridges the gap between generic skill templates and the nuanced realities of production codebases, such as specific framework patterns, naming conventions, and complex directory structures.
Improve Skill: The Meta-Skill Use Cases
- Refining a skill that recently produced suboptimal or generic results.
- Adapting an existing skill to work within a new project framework or monorepo.
- Identifying and fixing unhandled edge cases like binary files in source directories or specific encoding issues.
- Reducing token consumption by trimming unnecessary context in skill definitions.
- Automatically improving skills when the agent detects repeated manual corrections from the user.
How Improve Skill: The Meta-Skill Works
- The user triggers the evaluation by naming a specific skill or responding to an agent's suggestion to improve the last used tool.
- The skill performs a multi-point audit, starting with a definition check to ensure instructions are unambiguous.
- It scans the current project to identify file types, naming conventions, and directory structures that the target skill might miss.
- It evaluates prior output quality, checking for accuracy and consistency in generated artifacts.
- A structured Scorecard is generated, highlighting strengths and weaknesses across six key dimensions.
- Upon user approval, the skill generates and applies patches to the target SKILL.md file to implement the improvements.
Improve Skill: The Meta-Skill Setup
To use this meta-utility, ensure it is available in your skill library and invoke it via your agent interface. No complex configuration is required as it dynamically discovers installed Openclaw Skills.
/improve-skill [target-skill-name]
If no skill name is provided, the agent will list all available Openclaw Skills for you to choose from.
Improve Skill: The Meta-Skill Data Schema & Taxonomy
The skill generates a structured Improvement Report to document its findings. The data is organized into the following sections:
| Section | Description |
|---|---|
| Scorecard | A high-level rating (Strong to Missing) for clarity, fit, edge cases, quality, composability, and UX. |
| Top Improvements | A prioritized list of fixes categorized by impact (High/Medium) and effort (Trivial to Large). |
| Edge Case Inventory | A list of project-specific patterns and whether they are currently handled. |
| Suggested Patches | Specific Markdown additions or changes intended for the skill's definition file. |
Improve Skill: The Meta-Skill Advanced Features
- Context-aware discovery that automatically suggests improvements when it senses agent friction.
- Framework-specific intelligence that recognizes patterns in Next.js, Flask, and other modern stacks.
- Simulation mode that performs dry runs on representative files to predict output quality before making changes.
- Integration with other Openclaw Skills to ensure cross-skill hooks and data consistency are maintained during updates.
SKILL.md
Loading
Related Openclaw Skills
Ghost Catalog
An intelligent file management system that implements semantic headers and a SQLite database to organize, tag, and validate project assets.
Ghost Catalog
Ghost Catalog provides a semantic file header system and SQLite database to scan, tag, and organize files within any workspace for better discoverability and compliance.
Lobster Trap
A shared digital whiteboard for synchronized management of notes, code snippets, and key labels via API.
OpenDex API Skill
A high-performance interface for retrieving Solana market data and community-driven token insights.
ClawMeter
A self-hosted dashboard for monitoring OpenClaw API spending and token usage with real-time budget alerts.
Agent Chat
A specialized messaging protocol that enables multiple Openclaw Skills instances to coordinate and exchange data through a shared JSON queue.
Featured*
