A diagnostic tool within the Openclaw Skills ecosystem designed to audit LLM stacks against live OpenRouter pricing and performance benchmarks.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install model-audit
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 model-audit 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 Model Audit?
Model Audit is a professional-grade utility for developers and AI engineers who need to maintain a cost-effective and high-performance LLM stack. By integrating directly with the OpenRouter API, it provides a real-time snapshot of the market, comparing your current configurations against the latest models. This skill ensures your AI infrastructure remains on the cutting edge by identifying cheaper or more capable alternatives in categories such as reasoning, vision, and speed. Utilizing this tool as part of your Openclaw Skills library allows for automated stack optimization and significant reduction in operational overhead.
Model Audit Use Cases
- Periodic auditing of AI operational costs to discover budget-saving migration paths.
- Benchmarking internal model configurations against new market releases from providers like Anthropic and OpenAI.
- Comparing specific model pairs (e.g., Claude 3.5 Sonnet vs GPT-4o) for price-to-performance efficiency.
- Ensuring functional coverage across specialized categories like coding, vision, and low-latency inference.
How Model Audit Works
- The skill establishes a connection with the OpenRouter API to fetch the latest token pricing and model capabilities.
- It reads the local openclaw.json configuration to identify which models are currently being utilized by your agents.
- It classifies all models into technical categories such as reasoning, code, fast, cheap, and vision.
- The system performs a comparative analysis, calculating potential monthly savings based on current market rates.
- It generates a comprehensive report or a structured JSON object containing upgrade recommendations and cost-optimization strategies.
Model Audit Setup
To deploy this skill, ensure your environment is configured correctly:
# Ensure Python 3.10 or higher is installed
# Export your OpenRouter API Key to your environment
export OPENROUTER_API_KEY="your_api_key_here"
# Run a standard audit
python3 scripts/model_audit.py
# Run a comparative audit between specific models
python3 scripts/model_audit.py --compare "anthropic/claude-3.5-sonnet" "openai/gpt-4o"
Model Audit Data Schema & Taxonomy
The skill manages and generates data according to the following structure:
| Component | Details |
|---|---|
| Input Source | Reads active model IDs from openclaw.json |
| Market Data | Fetches live prompt and completion costs via OpenRouter API |
| Taxonomy | Categorizes models into: reasoning, code, fast, cheap, and vision |
| Output Modes | Supports human-readable CLI tables and machine-readable JSON formats |
Model Audit Advanced Features
- Automated CI/CD Integration: Use the
--jsonflag to feed audit data into automated monitoring or dashboarding tools. - Specific Model Targeting: Use the
--modelsflag to focus the audit on a subset of providers relevant to your specific project. - Benchmark Leaders: Use the
--topcommand to quickly identify the current price-to-performance leaders in every major LLM category. - Cost Projection: Automatically estimates potential savings based on your existing token consumption profile.
SKILL.md
Loading
Related Openclaw Skills
Model Audit
A comprehensive auditing tool that compares your current LLM stack against live OpenRouter pricing to find cheaper and more powerful model alternatives.
Mercury Banking
A professional-grade Mercury banking integration for AI agents that automates transaction tracking, cash flow analysis, and financial categorization.
Make (Integromat) Automation
A high-performance Python tool for managing Make automation scenarios, executions, connections, and data stores through a zero-dependency API interface.
Mac Mini AI Server for OpenClaw
A comprehensive deployment guide for running Openclaw Skills on a dedicated Mac Mini server for 24/7 AI availability.
Model Council
A multi-model consensus engine that queries several LLMs simultaneously and uses a judge model to synthesize the best possible answer.
Model Council
A multi-model consensus system that queries multiple LLMs simultaneously and uses a high-tier judge model to synthesize the best possible answer.
Featured*
