sudocode
Lightweight agent orchestration dev tool that lives in your repo
claude mcp add --transport stdio sudocode-ai-sudocode npx -y sudocode
How to use
sudocode is an MCP server that anchors and coordinates agent workflows within your repository. It exposes a local server/UI and a memory-backed coordination layer that helps you organize specs, issues, and agent execution trajectories. The CLI bundled with sudocode provides a familiar workflow for managing specs and issues, while the MCP server component enables real-time visualization, memory synchronization, and multi-agent orchestration. Use it to define high-level requirements as specs, map implementation plans as issues, and let agents execute topologically ordered workflows while recording their progress in a git-backed memory graph.
How to install
Prerequisites:
- Node.js and npm installed on your system
- Git installed (optional but recommended for repo-backed memory)
Installation steps:
- Install sudocode globally (or use npx to avoid version conflicts):
npm install -g sudocode
- Initialize a sudocode project in your repository (runs in your project directory):
sudocode init
- Start the local sudocode server (from anywhere in your project):
sudocode server
- Optional: connect agents to sudocode via the appropriate agent plugin instructions in your environment (e.g., Claude/Codex/Cursor plugins) as described in the quickstart guide: https://docs.sudocode.ai/quickstart
Additional notes
Tips and troubleshooting:
- If you encounter port or binding issues, check your environment for an existing running sudocode server and stop it before restarting.
- The MCP server component works best when paired with a git-backed repository, enabling Specs, Issues, and Workflows to be tracked as part of your codebase.
- Environment variables are typically minimal; primarily ensure Node.js is available and that your PATH includes the sudocode binary or npx access.
- For performance with large task graphs, consider increasing Node.js memory limits and ensuring your local machine has enough CPU cores to handle parallel agent execution.
- Refer to the Quick Start Guide for tutorials on setting up agent plugins and connecting external agents.
Related MCP Servers
systemprompt-code-orchestrator
MCP server for orchestrating AI coding agents (Claude Code CLI & Gemini CLI). Features task management, process execution, Git integration, and dynamic resource discovery. Full TypeScript implementation with Docker support and Cloudflare Tunnel integration.
guildbridge
🏰 Remotely hosted MCP server for Discord
devserver
A specialized MCP server for monitoring development server output with intelligent error categorization and Claude Code integration
mcp-jira-stdio
MCP server for Jira integration with stdio transport. Issue management, project tracking, and workflow automation via Model Context Protocol.
LinkedIn-Posts-Hunter
LinkedIn Posts Hunter MCP is a Model Context Protocol (MCP) server that provides tools for automating LinkedIn job post search and management through your AI assistant (Claude Desktop, Cursor, or other MCP-compatible clients).
docrag
AI-powered documentation RAG system with MCP server for Claude Code. Search and retrieve technical documentation on-demand with vector embeddings and smart web scraping.