Chapel Support for MCP

What is Chapel Support for MCP?

A Model‑Context‑Protocol (MCP) server for the Chapel programming language that provides tools for compiling, linting, and accessing educational primers, designed to integrate Chapel functionality with AI assistants and other tools.

How to use Chapel Support for MCP?

Install with UV, set the CHPL_HOME environment variable (or use a .env file or auto‑detection if chpl is in PATH), then run uv run chapel-support.py to start the server. Configure AI assistants to connect to it using a client configuration that points to the uv command and the server script.

Key features of Chapel Support for MCP

• Browse and access Chapel primer examples.
• Compile Chapel programs through the API.
• Lint Chapel code with chplcheck and auto‑fix.
• Automatically detect Chapel installation directory.
• List and enable custom linting rules.

Use cases of Chapel Support for MCP

• An AI assistant writing Chapel code can compile and check for errors on the fly.
• A learner can fetch primer contents to study Chapel’s parallel programming concepts.
• A developer can lint an existing Chapel codebase for style and best practices.
• Integrate Chapel development into AI‑driven workflows without leaving the chat interface.

FAQ from Chapel Support for MCP

What are the prerequisites to use Chapel Support for MCP?

Python 3.13 or higher and a working Chapel installation are required. For linting, chplcheck (shipped with Chapel) is optional.

How does Chapel Support for MCP find the Chapel installation?

It searches CHPL_HOME environment variable, then a .env file in the project root, then runs chpl --print-chpl-home if the compiler is in PATH.

What tools does Chapel Support for MCP offer?

It provides five tools: list_primers(), get_primer(path), compile_program(program_text, program_name), list_chapel_lint_rules(), and lint_chapel_code(program_text, program_name, fix, custom_rules).

How can I use Chapel Support for MCP with an AI assistant?

Configure the assistant to connect to the server using a client JSON object that specifies uv run --directory <path> chapel-support.py as the command.

Can Chapel Support for MCP automatically fix linting issues?

Yes. When fix=True is passed to lint_chapel_code(), the server applies automatic fixes from chplcheck and returns the corrected code.