#3434th of 64 in Agent and MCP starters
golf
Python framework that compiles tool, resource and prompt files into an MCP server
- Stars
- 843
- License
- Apache-2.0
- Last commit
- Sep 2026
- Last release
- Sep 2026
- Language
- Python
Missed the latest GitHub check; these facts may be older.
Overview
Golf builds an MCP server from a directory of Python files: each file in tools/, resources/ or prompts/ defines one component, and the component ID is derived from its path. It targets FastMCP 4.0.0 and MCP 2026-07-28, with legacy client compatibility. Authentication (JWT, OAuth server mode, static dev tokens) is configured in auth.py, and OpenTelemetry tracing is switched on through golf.json or environment variables.
Who it is for: Python developers building and self-hosting MCP servers
Strengths
- File-per-component layout; tool schemas inferred from function signatures and Pydantic models
- Auth options in auth.py: JWT with JWKS, OAuth 2.0 server mode, static dev tokens
- Optional OpenTelemetry tracing, with a detailed_tracing flag for input/output capture
- Supports streamable-http and stdio transports; targets MCP 2026-07-28 with legacy compatibility
Weaknesses
- Collects anonymous CLI usage telemetry by default; opt-out needed
- v0.2.0 auth API is a breaking change from v0.1.x
- Under MCP 2026-07-28, elicit and sample need InputRequiredResult return handling in tools
- SSE transport is deprecated; no Docker or compose files provided
What it needs
- Python
- Needs Python 3.10+, FastMCP 4.0.0
Also in Agent and MCP starters
See all 64| Rank | Project | Score |
|---|---|---|
| 1 | deepseek-harnessPlugin-based agent harness from DeepSeek with a local Web UI | 98 out of 100 |
| 2 | composioTypeScript and Python SDKs that give agents authenticated tools across 1000+ apps | 92 out of 100 |
| 3 | conductorDurable workflow engine that orchestrates microservices and LLM agents from JSON | 91 out of 100 |
| #4 | mcp-useTypeScript framework for building MCP servers with React Views and an Inspector | 87 out of 100 |
| #5 | a2uiProtocol and renderers for agents to stream native UI instead of text | 86 out of 100 |