An expert architect tool designed to analyze, refactor, and optimize AI agent skill definitions using industry-standard design patterns.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-optimizer-lilifeng
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-optimizer-lilifeng 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 Optimizer?
Skill Optimizer is a specialized utility within the Openclaw Skills ecosystem dedicated to the structural improvement of AI agent definitions. It serves as a bridge between raw agent instructions and high-performance, reliable logic by applying five core architectural patterns: Tool Wrapper, Generator, Reviewer, Inversion, and Pipeline. By analyzing SKILL.md files, it identifies logical flaws, removes hardcoded bottlenecks, and ensures that agents follow a robust Load-Process-Output workflow.
This skill is particularly valuable for developers working in the Google ADK ecosystem who need to standardize their agent behaviors. It transforms vague instructions into imperative, modular, and testable skill sets, significantly reducing hallucinations and improving the success rate of complex multi-stage tasks.
Skill Optimizer Use Cases
- Refactoring legacy SKILL.md files to improve execution reliability.
- Auditing new AI agents for structural flaws and logic gaps.
- Standardizing skill definitions across a large Openclaw Skills repository.
- Automating the optimization of skills immediately after file creation or modification.
- Enhancing agent performance when quality scores drop below defined thresholds.
How Skill Optimizer Works
- Intent Analysis: The skill parses the target SKILL.md file to identify the core mission and existing design patterns.
- Strategy Proposal: It generates a summary of current issues and proposes a specific optimization strategy involving modular references and pattern application.
- Structural Refactoring: Upon confirmation, it rewrites the file, moving long-form data to dynamic reference paths and enforcing strict instruction sets.
- Change Rationale: The skill provides a detailed changelog explaining pattern choices, context efficiency gains, and safety gate implementation.
- Validation Checklist: A final internal audit ensures the optimized skill meets all quality rubrics before deployment.
Skill Optimizer Setup
To integrate this optimizer into your Openclaw Skills workflow, ensure the skill-optimizer directory is present in your skills path. You can manually trigger an audit using the following command structure:
# Trigger optimization for a specific skill file
claw optimize skills/your-skill/SKILL.md
# Check quality of an existing agent
claw audit skills/your-skill/
Skill Optimizer Data Schema & Taxonomy
The Skill Optimizer organizes its analysis and output according to the following structure:
| Component | Description |
|---|---|
| Core Intent | The primary objective the skill must fulfill. |
| Design Patterns | Implementation of Tool Wrapper, Generator, Reviewer, Inversion, or Pipeline. |
| References | Externalized assets stored in references/ or assets/ folders. |
| Safety Gates | Explicit 'DO NOT' instructions to prevent model hallucination. |
| Pipeline Steps | A numbered sequence of execution checkpoints. |
Skill Optimizer Advanced Features
- Auto-Trigger Hooks: Automatically initiates an optimization scan when SKILL.md files are created or edited within a 3-minute window.
- Quality Scoring Integration: Suggests improvements automatically if the system detects a skill quality score below 80%.
- Modular Reference Extraction: Automatically identifies long-form text within a skill and moves it to hypothetical reference files for token efficiency.
- Pattern-Based Enforcement: Dynamically switches its refactoring logic based on whether the agent is a Generator (output-focused) or a Reviewer (audit-focused).
SKILL.md
Loading
Related Openclaw Skills
Podcastfy (Clawdbot)
A powerful AI skill that converts web content, PDFs, and articles into natural-sounding audio conversations using Gemini and Edge TTS.
ADB Organization Info Assistant
An intelligent information assistant providing real-time data, reports, and organizational insights regarding the Asian Development Bank (ADB).
NOC Engineer Roadmap
A career development platform that generates structured, personalized learning paths for aspiring Network Operations Center engineers.
Shopify Social Content Planner
Automate your e-commerce marketing with a data-driven 30-day social media calendar designed specifically for Shopify store growth.
ObjectRemover Video Object Ops
An AI-powered video editing skill for removing unwanted objects or extracting subjects using natural language or API workflows.
Acura Automotive Info Assistant
An intelligent information assistant providing real-time data on Acura vehicle lineups, technical specifications, and dealership details.
Featured*
