relayos-mcp
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
relayos-mcp
🌟 核心理念:以最小 Token 消耗,做最高质量的交付
graph LR
User["👤 开发者需求"] --> Orchestrator["🧠 Claude Code / Cursor (统帅)"]
Orchestrator -->|"MCP 调度协议"| RelayOS["⚡ relayos-mcp 核心网关"]
RelayOS -->|"免费并行派发"| Worker1["💻 Mimo AI (基础分析)"]
RelayOS -->|"并行任务"| Worker2["🛠️ OpenCode (代码生成)"]
RelayOS -->|"本地沙箱执行"| Worker3["⚙️ Local CLI (测试与运行)"]
Worker1 --> RelayOS
Worker2 --> RelayOS
Worker3 --> RelayOS
RelayOS -->|"聚合报告"| Orchestrator
Orchestrator -->|"高质量收敛输出"| User💰 降本 90%:单调重复的文档抓取、语法检查、单元测试跑批全部下沉给免费或轻量终端 Agent;
⚡ 并行异步:多任务并发调度,不阻塞主对话上下文;
🎯 质量把关:Claude Code 专注于核心架构设计、逻辑审核与最终代码合入。
Related MCP server: llm-router-mcp
🚀 快速安装
1. 安装 pip 包
pip install relayos-mcp[mcp]2. 检查本地 Agent 可用性
relay-task agents输出应显示可用 Agent 列表:
[ok] mimo (local / free)
[ok] opencode (terminal worker)
[ok] claude (deep reasoning)3. 配置到 Claude Code
编辑 ~/.claude/settings.json(或项目级 .claude.json):
{
"mcpServers": {
"relayos": {
"command": "relayos-mcp",
"args": []
}
}
}如果你使用 Claude Desktop,同样可将上述配置添加至 claude_desktop_config.json。
🛠️ MCP 工具矩阵
工具名称 | 对应功能 | 推荐场景 | 计费成本 |
| 本地安全沙箱命令执行 | 跑测试、构建项目、环境探测 | 免费 |
| 调度 Mimo AI 终端 | 基础语法审查、初阶重构建议 | 免费 |
| 调度 OpenCode 编写代码 | 批量模版生成、脚手架构建 | 免费 |
| 复杂深度多步推理任务 | 高难度算法推演、架构最终仲裁 | 按 Token 计费 |
💡 典型实战场景
你: "给这个项目写完全部单元测试,并跑通 pytest"
└─► Claude Code (主脑规划):
├─► 调用 ask_opencode: 生成 5 个模块的基础单元测试代码 (免费)
├─► 调用 run_command: "pytest tests/ -v" (免费本地执行)
├─► 发现 1 个边界条件断言失败
└─► Claude Code 亲自修复 bug,提交代码,输出优雅报告给你!📂 项目结构
relayos-mcp/
├── relayos_mcp/
│ ├── __init__.py
│ ├── server.py # MCP 服务端核心实现
│ ├── agents/ # Mimo / OpenCode / Claude 适配驱动
│ └── utils/ # 协议编解码与超时重试控制
├── claude_prompt.md # 推荐给 Claude Code 的系统提示词模板
├── pyproject.toml # 构建配置文件
└── README.md📄 开源许可证
本项目采用 MIT 许可证。欢迎提交 Issue 或 Pull Request!
This server cannot be deployed
Maintenance
Related MCP Connectors
Coordinate coding agents through MCP using existing AI plans, saved work, and independent checks.
AI work orchestration for plans, tasks, teams, and coding-agent dispatch.
Use AI models for chat, image, and video generation from Claude Code and other MCP hosts.
One MCP endpoint for Claude, GPT & Gemini: 100+ tools + no-code connectors + agent workers.
Related MCP Servers
- AlicenseBqualityFmaintenanceEnables orchestrating multiple AI CLI agents (Claude Code, Codex, Gemini CLI, Copilot CLI) through a unified MCP interface for task delegation, cross-agent comparison, and specialized tools like code review and debugging.149 npm14MIT
- AlicenseDqualityCmaintenanceEnables Claude Code as a team leader to delegate tasks to GPT and Gemini experts, supporting multi-LLM collaboration with tools for consultation, code review, design, and background execution.7624 npmMIT
- AlicenseAqualityDmaintenanceAn MCP server that enables Claude to orchestrate multiple autonomous Claude Code agents working in parallel across different projects, with tools to dispatch, monitor, and manage their progress.637 npm1MIT
- AlicenseNot gradedqualityAmaintenanceEnables multi-agent AI collaboration by connecting Claude Code, OpenAI Codex, and Google Antigravity, providing cross-model code reviews, independent second opinions, and autonomous validation through MCP tools.1MIT