An automated diagnostic tool that reviews AI agent configuration files to provide scoring and optimization insights.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install claw-score
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 claw-score 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 Claw Score Agent Architecture Audit?
Claw Score is a comprehensive auditing utility built for the Openclaw Skills ecosystem. It allows developers to submit their agent's core architectural files—such as instructions, personality profiles, and security rules—for a professional review by Atlas. The skill ensures that your agent follows best practices in agentic design while maintaining strict privacy standards.
By leveraging this tool, you can transform raw configuration files into high-performing digital entities. It bridges the gap between basic setup and advanced agent engineering, making it an essential addition to any developer's toolkit of Openclaw Skills.
Claw Score Agent Architecture Audit Use Cases
- Auditing agent logic for better reliability before production deployment.
- Finding and removing accidental PII, credentials, or API keys from workspace files.
- Benchmarking agents against the Mega Claw scoring tier to ensure high quality.
- Getting actionable implementation examples for agent improvement and quick wins.
How Claw Score Agent Architecture Audit Works
- The skill identifies architectural files within the workspace such as AGENTS.md and SOUL.md.
- Automated sanitization logic strips API keys, email addresses, and other personal identifiers.
- The user reviews a preview of the sanitized data to confirm the submission.
- Files are transmitted securely to Atlas for professional architectural analysis.
- A detailed report containing a score from 1 to 5 and specific recommendations is delivered via email.
Claw Score Agent Architecture Audit Setup
Install the audit skill directly into your workspace to start using it with other Openclaw Skills:
# If using OpenClaw skill system
cp -r /path/to/claw-score skills/
# Or download from ClawhHub
npx clawhub install claw-score
Claw Score Agent Architecture Audit Data Schema & Taxonomy
The skill processes a specific set of metadata files to generate the Claw Score:
| File | Description |
|---|---|
| AGENTS.md | Main agent instructions and logic |
| SOUL.md | Personality and identity configuration |
| MEMORY.md | Long-term memory and context settings |
| TOOLS.md | Capability and tool definitions |
| SECURITY.md | Safety rules and constraint logic |
Claw Score Agent Architecture Audit Advanced Features
- Automated multi-pattern credential stripping for sk-, xoxb-, and environment variables.
- Comprehensive workspace file tree analysis for structural auditing.
- Detailed scoring across six distinct technical dimensions.
- Actionable recommendations featuring copy-paste implementation examples.
SKILL.md
Loading
Related Openclaw Skills
EVM Wallet Skill for AI Agents
A self-sovereign EVM wallet enabling AI agents to manage assets, swap tokens, and access private AI inference via Venice AI.
Molt Life Kernel
A specialized Node.js infrastructure module providing AI agents with persistent identity, crash recovery, and cognitive health monitoring.
molt-life-kernel
A specialized Node.js infrastructure module providing AI agents with persistent identity, crash recovery, and cognitive health monitoring through a verifiable ledger system.
Sphero Mini Control
A Python-based automation skill to control Sphero Mini robot balls through Bluetooth Low Energy across multiple platforms.
OpenClaw Command Center
A mission control dashboard providing real-time session monitoring and cost intelligence for your AI agent workforce.
Spacesuit Framework for OpenClaw
A batteries-included framework scaffold designed to standardize and secure AI agent workspaces.
Featured*
