What are Specialist Roles in Owlex?

Specialist Roles allow you to assign specific perspectives to agents, such as 'security' to find vulnerabilities, 'perf' for resource usage, or 'skeptic' to identify reasoning flaws.

How do I install the Owlex skill?

You can install it using the command 'uv tool install git+https://github.com/agentic-mcp-tools/owlex.git' and then adding the server to your .mcp.json configuration.

Which LLMs can be used with Owlex?

Owlex supports various models including Codex, Gemini, and OpenCode, allowing you to leverage the unique strengths of different AI providers.

What is the Owlex AI Council?

Owlex is a multi-agent deliberation server that allows multiple LLMs to critique and refine ideas in a structured, multi-round process directly from your CLI.

How does the council deliberation process work?

It follows a three-step process: agents first answer independently, then review and revise based on each other's input, and finally, a primary agent synthesizes the debate into a recommendation.

Owlex AI Council

Name: Owlex AI Council
Author: robertpelloni

byrobertpelloni

•

Data Science & ML

Orchestrates multi-agent deliberations and structured critiques across multiple LLMs to provide optimized architectural and coding recommendations.

Owlex acts as a sophisticated deliberation server that brings the power of multiple AI models—such as Codex, Gemini, and OpenCode—into a single CLI workflow. It operates in distinct phases—independent reasoning, cross-agent deliberation, and final synthesis—to ensure complex technical decisions are vetted through specialized perspectives like security, performance, and architectural integrity. This skill is invaluable for developers seeking high-confidence 'second opinions' on critical code changes, complex bug fixes, or large-scale system design problems where a single model's perspective might be limited.

Key Features

01Cross-model synthesis to consolidate diverse AI insights

023 GitHub stars

03Specialized persona roles including Security, Performance, and Architect

04Persistent individual sessions with specific LLM engines

05Team presets for targeted audits and architecture reviews

06Multi-agent deliberation rounds for objective consensus

Use Cases

01Vetting complex architectural decisions for microservices or monorepos

02Performing deep security audits using the devil's advocate perspective

03Optimizing code performance through multi-model peer review and critique

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add robertpelloni/borg owlex

For use in Claude.ai and ChatGPT

Download Skill