Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "thug-fugu": {
      "command": "uvx",
      "args": [
        "thug-fugu"
      ]
    }
  }
}

Or install with pip: pip install thug-fugu

README Excerpt

<img width="1254" height="1254" alt="Thug-Fug-Logo" src="docs/assets/Thug-Fug-Logo.png" /> Thug AI の Fugu のように、複数ロールのローカル LLM を協調実行するための最小 Python 実装です。 > **English:** A minimal Python toolkit with a standard-library-only core for orchestrating **multiple local LLM roles** (planner / coder / reviewer / synthesizer) in parallel. It talks to **Ollama** and **OpenAI-compatible servers** (LM Studio, llama.cpp server, vLLM, …), merges the roles into a single answer, and exposes an **OpenAI Chat Comple

Tools (20)

always_includeapi_keybackendbase_urlcoordinatordirectechois_synthesizeris_verifierkeywordsmax_parallel_workersmax_tokensmodelmodel_poolsollamaparallel_ensemblepolicyrequest_timeout_secondsrole_splitselection_policy

Thug-Fugu

Quick Install

README Excerpt

Tools (20)

Related AI/LLM Servers