This MCP server entry points to the MCPStack Tool Builder, a Python-based template to accelerate creating MCPStack tools. It provides a scaffolding workflow and a CLI (mcpstack_tool.py) to bootstrap, preview, apply, and validate tool configurations. The server enables you to generate a tool skeleton, customise your tool slug, class name, and environment prefixes, and then integrate the tool into an MCPStack pipeline. Use the CLI through Python (or via UV if you have UV installed) to access commands such as init, preview, apply, and validate. When running, you can invoke the server by executing the configured command and arguments in your environment, for example: python mcpstack_tool.py --help to see available options.

Prerequisites:

• Python 3.9+ and a working virtual environment
• Git
• Optional: UV (recommended for development workflows)

Step-by-step installation:

1. Clone the repository: git clone https://github.com/MCP-Pipeline/MCPStack-Tool-Builder.git cd MCPStack-Tool-Builder

2. Install dependencies (development mode): pip install -e .[dev]

3. (Optional) Set up development tooling (pre-commit): uv run pre-commit install

   or legacy option

   pre-commit install

4. Verify installation by listing available commands: python mcpstack_tool.py --help

5. Run the tool workflow via the MCPStack Tool Builder entry (as configured by the MCP config): python mcpstack_tool.py init # start interactive skeleton creation

Prerequisites recap: ensure Python 3.9+ is installed, you have Git access, and you have installed the project in editable/development mode to pick up changes while developing.

Tips and notes:

• This template focuses on quickly scaffolding an MCPStack tool with a CLI that supports init, preview, apply, and validate workflows. Use the init flow to populate tool_slug, class_name, and env_prefix.
• If you prefer running tools via UV, you can adapt the invocation to uv run mcpstack_tool.py or uv run <script> depending on your environment.
• Environment variables can influence tool generation (e.g., MCPSTACK_TOOL_BUILDER_MODE for debugging). Add or adjust env vars in the MCP config as needed.
• When integrating the generated tool into an MCPStack pipeline, remove unnecessary boilerplate to tailor the tool to your specific model/context and ensure you follow the MCPStack documentation for tool interfaces.
• Common issues: missing dependencies, mismatched Python versions, or placeholders not replaced during Apply. Use the Validate command to detect placeholders remaining after generation.