An automated diagnostic tool that analyzes AI conversation logs to evaluate performance and generate data-driven optimizations for agent skills.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install skill-optimizer
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 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?
The Skill Optimizer is a sophisticated development tool designed to bridge the gap between a skill's intended design and its actual execution. By processing real-world interaction logs, it provides developers with a deep-dive analysis into how an agent interprets instructions and where it might be failing to meet user expectations. This tool is a critical asset for anyone looking to maintain high-quality Openclaw Skills by providing a feedback loop that transforms raw logs into structured improvement plans.
It focuses on five core dimensions: trigger accuracy, workflow adherence, output quality, user interaction flow, and instruction effectiveness. By synthesizing these metrics, it ensures that your AI remains sharp, relevant, and efficient, moving away from guesswork and toward a systematic approach to agent refinement.
Skill Optimizer Use Cases
- Analyzing specific AI conversation logs to determine why a skill failed to trigger or behaved unexpectedly.
- Refining the metadata and description of Openclaw Skills to improve intent matching and reduce false activations.
- Identifying bottlenecks or redundant steps in complex multi-phase agent workflows.
- Iteratively improving the prompt engineering within a skill file based on evidence of instruction neglect by the LLM.
How Skill Optimizer Works
- The user provides a text-based conversation log or a file path containing the interaction history between a user and an AI.
- The optimizer identifies the specific Openclaw Skills mentioned in the logs or asks the user for clarification.
- It reads the raw SKILL.md documentation to compare the intended workflow against the actual conversation performance.
- A comprehensive analysis report is generated, covering intent matching, workflow execution, and output quality.
- The tool presents prioritized optimization suggestions with specific "before and after" changes for the markdown configuration.
- Upon user confirmation, it automatically patches the original SKILL.md file with the improved logic.
Skill Optimizer Setup
To get started with the Skill Optimizer, ensure you have the relevant conversation logs accessible to your agent environment. No complex installation is required beyond adding the skill to your library.
# Invoke the optimizer by providing a log file
openclaw run skill-optimizer --input ./logs/conversation.txt
The skill will automatically request permissions to read and write to your local Openclaw Skills directory to perform the optimization analysis.
Skill Optimizer Data Schema & Taxonomy
The skill processes log data and skill definitions to produce a structured analysis report. The output follows this taxonomy:
| Component | Description |
|---|---|
| Basic Info | Captures skill name, version, and the number of conversation rounds analyzed. |
| Dimensional Analysis | Detailed breakdown of trigger accuracy, workflow adherence, and instruction effectiveness. |
| Suggestion Priority | A ranked list of proposed changes to the SKILL.md file. |
| Change Summary | A diff-style view of proposed modifications including the rationale for each update. |
Skill Optimizer Advanced Features
- Intelligent intent inference that identifies which Openclaw Skills were meant to be active even if they weren't triggered.
- Automated file patching that directly updates skill documentation while preserving existing formatting.
- Multi-dimensional scoring that evaluates the "Interaction Experience" to ensure the agent's tone remains professional and helpful.
- Iterative improvement tracking to prevent over-optimizing based on a single, isolated interaction.
SKILL.md
Loading
Related Openclaw Skills
WeChat Moments Copywriting Guide
A conversational AI guide designed to help users rediscover their authentic voice and write personalized WeChat Moments posts through guided reflection.
WeChat Moments Assistant (Pangyu)
A personalized AI copywriting assistant that learns your unique voice to create authentic WeChat Moments posts.
Blockbuster Course Consultant (Dapianke)
A professional course design consultant that implements the Blockbuster Course methodology to guide users through the entire curriculum development lifecycle.
Stripe SQL Data Access
Direct SQL access to your Stripe billing and customer data via a synced PostgreSQL database.
MemoryBox
MemoryBox is a zero-dependency CLI tool designed to maintain a lean 3-tier hierarchical memory structure for Openclaw Skills.
OpenClaw Self-Healing System
An autonomous 4-tier recovery system that monitors gateway health and uses AI to diagnose and repair persistent failures.
Featured*
