Pyxel MCP
Let your agent scaffold and edit Pyxel retro Python games through MCP while you build a small title or jam entry.
Overview
Pyxel MCP is a MCP server for the Build phase that exposes the Pyxel retro Python game engine to coding agents over stdio.
What is this MCP server?
- MCP bridge for Pyxel, the Python retro game engine
- PyPI package pyxel-mcp v0.9.2 with stdio transport
- GitHub source at kitao/pyxel-mcp for engine-aligned tooling
- Suited to game jams, educational projects, and micro-indie titles
- Keeps sprite, tile, and game-loop edits inside agent-assisted sessions
- MCP server version 0.9.2 (PyPI pyxel-mcp)
- Transport: stdio
- Targets Pyxel retro game engine for Python
Community signal: 19 GitHub stars.
What problem does it solve?
Building Pyxel games with an AI helper means pasting engine docs and sample code instead of callable, structured tools.
Who is it for?
Indie devs and learners shipping small retro Python games who already chose Pyxel and want MCP in Claude Code or Cursor.
Skip if: 3D or Unity/Godot pipelines that do not use the Pyxel engine.
What do I get? / Deliverables
After install, your agent can interact with Pyxel-oriented MCP tools so game files and engine calls stay coordinated in one workflow.
- Agent-accessible Pyxel engine workflow during game builds
- Faster iteration on Pyxel scripts and assets from chat
- Documented MCP server entry for local game projects
Recommended MCP Servers
Journey fit
How it compares
Engine-specific MCP for Pyxel games, not a cross-engine game-design methodology skill.
Common Questions / FAQ
Who is Pyxel MCP for?
Python developers and jam participants building retro games with Pyxel who want engine-aware help from MCP-enabled agents.
When should I use Pyxel MCP?
Use it during Build frontend/game implementation when you are actively coding Pyxel projects and want tool calls instead of manual API lookup.
How do I add Pyxel MCP to my agent?
Install pyxel-mcp from PyPI (0.9.2), configure stdio MCP in your host, ensure Pyxel is available in the same Python environment, and restart the agent.