A multi-model consensus engine that queries several LLMs simultaneously and uses a judge model to synthesize the best possible answer.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install model-council-pro
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-council-pro 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 Council?
Model Council is a sophisticated multi-agent orchestration tool designed to eliminate the single-point-of-failure risk associated with relying on a single AI model. By leveraging OpenRouter, it distributes a single query across three or more distinct large language models (LLMs), such as Claude, GPT-4, and Gemini. This provides a diverse range of perspectives and technical approaches to any given problem.
This skill acts as a board of AI advisors, ensuring that mission-critical tasks are cross-verified by different architectures. Once the responses are gathered, a high-tier judge model evaluates the outputs, selects a winner based on reasoning, and provides a synthesized final response. This skill is a vital addition to any toolkit involving Openclaw Skills, ensuring high-fidelity outputs for complex decision-making and research.
Model Council Use Cases
- Making complex architectural or business decisions by gathering diverse AI perspectives.
- Performing deep code reviews to identify edge cases that a single model might overlook.
- Verifying research facts and data by cross-referencing information across different training sets.
- Comparing creative writing styles to select the most appropriate tone for specific audiences.
- Debugging stubborn software errors where different models might offer unique troubleshooting paths using Openclaw Skills.
How Model Council Works
- The user submits a query along with optional parameters for specific models and a designated judge.
- The system broadcasts the query to at least three different LLMs simultaneously via the OpenRouter API.
- Each council member model generates its independent response and returns it to the orchestrator.
- The designated judge model receives all responses, evaluates their logic, accuracy, and depth.
- The judge produces a final report including a winner, detailed reasoning, a synthesized answer, and a total cost breakdown.
Model Council Setup
To use this within your Openclaw Skills environment, ensure you have an OpenRouter API key and Python 3.10+ installed.
export OPENROUTER_API_KEY='your_api_key_here'
# Basic execution
python3 scripts/model_council.py "What's the best database for a real-time analytics dashboard?"
You can also customize the models used in the council:
python3 scripts/model_council.py --models "anthropic/claude-3.5-sonnet,openai/gpt-4o" "Your question"
Model Council Data Schema & Taxonomy
The skill processes inputs via CLI arguments and generates formatted console outputs or structured JSON for integration.
| Component | Description |
|---|---|
| council_members | List of LLMs queried (Default: Sonnet 3.5, GPT-4o, Gemini 2.0) |
| judge_model | The high-reasoning model that evaluates responses |
| reasoning | Markdown string explaining why a specific model won |
| synthesis | The final combined and verified answer |
| cost_metadata | Breakdown of API costs per model and total session cost |
Model Council Advanced Features
- Customizable council composition allows users to select specific models for niche expertise.
- Flexible judge selection enables high-tier models to oversee lower-latency models for cost-effective accuracy.
- JSON output mode available via the
--jsonflag for easy integration into automated pipelines and other Openclaw Skills. - Adjustable token limits and timeouts to manage performance for large-scale research queries.
- Real-time cost tracking to monitor API expenditure across different providers.
SKILL.md
Loading
Related Openclaw Skills
Model Audit
A diagnostic tool within the Openclaw Skills ecosystem designed to audit LLM stacks against live OpenRouter pricing and performance benchmarks.
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.
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.
Model Intel
A powerful tool for fetching live LLM pricing, capabilities, and performance comparisons directly from OpenRouter.
Featured*
