gonzalorando/rep-pilot repository overview

⁠Rep-Pilot

The AI Resources Hub — A centralized platform to discover, curate, and share agents, skills, instructions, rules, and MCP servers for AI-powered coding assistants.

🌐 reppilot.com⁠

⁠What is Rep-Pilot?

Rep-Pilot is a corporate repository for AI development resources. It acts as a centralized hub where every developer in your organization can publish their agents, skills, instructions, rules, and MCP servers — making them instantly discoverable and reusable by the entire team. Developers feed off each other's work, eliminating silos and keeping everything globalized across the company.

With Rep-Pilot you can:

• Publish and discover — each developer contributes their best resources to the shared catalog. Everyone browses, searches by tags, and stars what they find useful. No more reinventing the wheel.
• Scan your Git repositories (GitHub & GitLab) and auto-detect AI resources are recognized instantly.
• Generate personalized kits — describe your project needs, answer a few questions, and the built-in LLM engine selects the most relevant resources from the corporate catalog, downloads them, and packages a .zip kit ready for your AI coding tool (GitHub Copilot, Claude Code, OpenCode, and more).
• Chat with the LLM to refine your setup and get contextual guidance.

It integrates with your existing enterprise infrastructure — LDAP for authentication, any OpenAI-compatible LLM provider, and both GitHub and GitLab APIs.

⁠Quick Start

⁠Docker Compose

docker compose up -d

Open http://localhost⁠ and log in with admin / the password you set in ADMIN_PASSWD.

services:
  # MongoDB 7.0 — DataBase
  mongo:
    image: mongo:7.0
    container_name: rep-pilot-mongo
    restart: unless-stopped
    ports:
      - "27017:27017"
    volumes:
      - mongo-data:/data/db

  # Rep-Pilot — frontend + backend
  rep-pilot:
    image: gonzalorando/rep-pilot:latest
    container_name: rep-pilot
    restart: unless-stopped
    ports:
      - "80:80"
    environment:
      MONGO_URI: "mongodb://mongo:27017/rep-pilot"
      JWT_SECRET: your-secure-jwt-secret
      ADMIN_PASSWD: your-admin-password
      CORS_ORIGIN: "*"
      DEFAULT_LANGUAGE: "en"
      DATA_DIR: "/app/back/data"
    volumes:
      - rep-pilot-data:/app/back/data
    depends_on:
      mongo:
        condition: service_healthy

volumes:
  mongo-data:
  rep-pilot-data:

⁠Docker Run (rep-pilot only)

docker run -d \
  -p 80:80 \
  -e MONGO_URI=mongodb://host.docker.internal:27017/rep-pilot \
  -e JWT_SECRET=your-secure-jwt-secret \
  -e ADMIN_PASSWD=your-admin-password \
  -e CORS_ORIGIN=* \
  -e DEFAULT_LANGUAGE=en \
  -v rep-pilot-data:/app/back/data \
  --name rep-pilot \
  gonzalorando/rep-pilot:latest

⁠Environment Variables

⁠Volumes

⁠How to Use

1. Log in at http://localhost with admin / your password.
2. Configure your LLM provider (OpenAI-compatible), Git instances (GitHub/GitLab tokens), and optionally LDAP in the Admin panel.
3. Scan repositories — paste a Git URL and Rep-Pilot auto-detects agents, skills, instructions, and MCP servers from the file tree.
4. Browse the catalog — explore resources by name, type, or tags. Star your favorites.
5. Generate a kit — go to the IA Kit page, describe your project (e.g. "microservices with Spring Boot"), answer the interactive questions, and download a ready-to-use .zip structured for your AI tool of choice.