turbo-flow-claude

Turbo-Flow Claude is an integrated MCP server that brings together Claude Flow orchestration, AI agent capabilities, and developer tooling to accelerate end-to-end AI-assisted development. It exposes a suite of tools including Claude Code for AI-assisted coding, Claude Flow for orchestrating agent workflows, Claudish as a multi-model proxy, agtrace for agent observability with a live dashboard, and support for Spec-Kit and OpenSpec workflows. You can start the server locally and use its built-in MCPs to manage agent tasks, run cross-model analyses, and validate specs as you iterate on projects. The environment is designed to work in DevPod, GitHub Codespaces, or Google Cloud Shell, enabling rapid prototyping and collaboration. To interact with the server, run the provided start command and then use the included aliases and commands to launch specific capabilities like spec-driven development, multi-model coordination, and agent observation.

Prerequisites:

• Node.js (v14+ recommended) and npm
• Git
• Access to required API keys (OpenAI/Claude Flow/PAL) if you plan to enable model-based features

1. Clone the repository or pull the Turbo-Flow Claude project into your workspace

2. Install dependencies

npm install

3. Configure environment variables

• Create a .env file or set environment variables for:
  • OPENAI_API_KEY (or equivalent Claude Flow API key)
  • CLAUDE_FLOW_API_KEY (if using Claude Flow integrations)
  • PAL_API_KEY (if you enable PAL multi-model orchestration)

4. Start the MCP server

node server.js

5. Verify the server is running by visiting the appropriate local URL or checking logs for startup confirmation.

Tips:

• Ensure API keys are kept secure and not committed to version control.
• If you encounter initialization or path issues, check that you are running from the workspace root and that all dependencies are installed.
• The server integrates multiple tools (Spec-Kit, OpenSpec, agtrace, Claudish, PAL). Familiarize yourself with their specific commands and aliases in the README to maximize productivity.
• For debugging, use agtrace to monitor agent sessions and claude mcp commands to query or extend MCP capabilities.

Common issues:

• Missing API keys or invalid keys can cause model initialization failures. Double-check env vars.
• Dependency conflicts after npm install can be resolved by removing node_modules and reinstalling.
• If the server cannot access external models, verify network egress and API access permissions.