Server
Provides a robust backend for AI agents, delivering custom instruction files and executable scripts.
This Python-based MCP (Multi-Cloud Platform) server, built with FastAPI and Uvicorn, acts as a central hub for managing and serving custom AI agent instructions and bash scripts. It maintains an organized 'agents-library' containing AGENTS.md files for agent rules, prompts, and contextual information, alongside executable bash scripts. Designed for seamless integration, it enables AI command-line interfaces like `gemini-cli` to retrieve specific agent instructions and execute tools, enhancing AI's capability to understand context and perform actions.
主な機能
01Serves custom AGENTS.md instruction files for AI agents
02Supports containerization with Docker and Docker Compose
03Hosts and executes custom bash scripts via API
04Integrates with `gemini-cli` for AI agent context and tool invocation
050 GitHub stars
06Built with FastAPI for high-performance API delivery
ユースケース
01Enabling consistent AI agent behavior across development and deployment environments
02Providing dynamic, up-to-date agent prompts and scripts to AI-powered CLIs
03Centralizing and managing AI agent instructions and contextual rules