metacoder
Wrapper for multiple AI coders (goose, claude code, codex)
claude mcp add --transport stdio ai4curation-metacoder npx -y metacoder
How to use
Metacoder provides a single CLI interface to multiple AI coding assistants (such as goose, claude, codex, gemini, qwen) and supports working with MCP collections. You can run Metacoder to generate code or refactor snippets using your chosen coder, switch coders on the fly, and manage MCP-based collections of tasks. The tool also supports using MCPs (GitHub MCPs or other MCP collections) to run more complex workflows, like applying a predefined set of coding tasks against multiple coders or models. Typical workflows include selecting a coder with -c, pointing Metacoder at a working directory with -w, and optionally supplying an MCP collection file with --mcp-collection to apply a predefined MCP across coders and models. The interface is designed to be consistent across coders and to make evaluating different agents and MCPs easier in pipelines and local runs.
How to install
Prerequisites:\n- Node.js and npm installed on your system (Metacoder is commonly distributed as a Node-based CLI).\n- Basic familiarity with running commands in a terminal.\n\nInstallation steps:\n1) Install Node.js from https://nodejs.org/ (npm is included with Node.js).\n2) Install Metacoder globally or use npx to run it directly:\n\n - Global install (optional): npm install -g metacoder\n - Or run via npx without global install: npx -y metacoder <args>\n3) Verify installation: metacoder --help or npx -y metacoder --help.\n\nUsage example after installation:\n- List available coders: metacoder list-coders\n- Run with a specific coder: metacoder "Write a Python function to calculate fibonacci numbers" -c claude -w my-scripts/\n- Use MCP collections: metacoder "Fix issue 1234" -w path/to/my-repo --mcp-collection github_mcps.yaml
Additional notes
Tips and common considerations:\n- Metacoder supports multiple coders with a unified CLI, plus the ability to apply MCPs for more complex workflows.\n- MCPs allow orchestration across coders/models and can be provided as YAML/JSON collections.\n- When using MCPs, you can specify environment variables needed by the MCP (for example, credentials or endpoints) via the env section of the MCP configuration.\n- Ensure your working directory (-w) is writable and contains the files you intend to process.\n- If you encounter issues with a specific coder, check the coder’s documentation or update to the latest Metacoder release.
Related MCP Servers
wanaku
Wanaku MCP Router
sympy
A MCP server for symbolic manipulation of mathematical expressions
okta
The Okta MCP Server is a groundbreaking tool built by the team at Fctr that enables AI models to interact directly with your Okta environment using the Model Context Protocol (MCP). Built specifically for IAM engineers, security teams, and Okta administrators, it implements the MCP specification to help work with Okta enitities
mcp-manager
CLI tool for managing Model Context Protocol (MCP) servers in one place & using them across them different clients
mcp-oi-wiki
🌟 Wiki of OI / ICPC for LLMs. (面向大模型的某大型游戏线上攻略,内含炫酷算术魔法)
linkedin_mcp
A powerful Model Context Protocol server for LinkedIn interactions that enables AI assistants to search for jobs, generate resumes and cover letters, and manage job applications programmatically.