vrchat-udon-mcp
```
██╗ ██╗██████╗ ██████╗██╗ ██╗ █████╗ ████████╗ ██╗ ██╗██████╗ ██████╗ ███╗ ██╗ ███╗ ███╗ ██████╗██████╗
██║ ██║██╔══██╗██╔════╝██║ ██║██╔══██╗╚══██╔══╝ ██║ ██║██╔══██╗██╔═══██╗████╗ ██║ ████╗ ████║██╔════╝██╔══██╗
██║ ██║██████╔╝██║ ███████║███████║ ██║ ██║ ██║██║ ██║██║ ██║██╔██╗ ██║ ██╔████╔██║██║ ██████╔╝
██║ ██║██╔══██╗██║ ██╔══██║██╔══██║ ██║ ██║ ██║██║ ██║██║ ██║██║╚██╗██║ ██║╚██╔╝██║██║ ██╔═══╝
╚██████╔╝██║ ██║╚██████╗██║ ██║██║ ██║ ██║ ╚██████╔╝██████╔╝╚██████╔╝██║ ╚████║ ██║ ╚═╝ ██║╚██████╗██║
╚═════╝ ╚═╝ ╚═╝ ╚═════╝╚═╝ ╚═╝╚═╝ ╚═╝ ╚═╝ ╚═════╝ ╚═════╝ ╚═════╝ ╚═╝ ╚═══╝ ╚═╝ ╚═╝ ╚═════╝╚═╝
```
# VRChat Udon MCP
[](https://github.com/MauDevVR/vrchat-udon-mcp/releases)
[](LICENSE)
[](https://nodejs.org/)
[](https://modelcontextprotocol.io)
**MCP server for VRChat UdonSharp development** — exposes the [agent-skills-vrc-udon](https://github.com/niaka3dayo/agent-skills-vrc-udon) knowledge base to AI assistants via the [Model Context Protocol](https://modelcontextprotocol.io).
---
## 🌐 Documentation
| Language | README | Description |
|----------|--------|-------------|
| 🇺🇸 **English** | [docs/README.en.md](docs/README.en.md) | Full guide — installation, MCP setup, tools |
| 🇪🇸 **Español** | [docs/README.es.md](docs/README.es.md) | Guía completa en español |
| 🇫🇷 **Français** | [docs/README.fr.md](docs/README.fr.md) | Guide complet en français |
| 🇯🇵 **日本語** | [docs/README.ja.md](docs/README.ja.md) | 日本語の完全ガイド |
| 🇰🇷 **한국어** | [docs/README.ko.md](docs/README.ko.md) | 한국어 전체 가이드 |
---
## What is this?
`vrchat-udon-mcp` is a **stdio MCP server** that indexes, searches, and validates UdonSharp documentation from the `agent-skills-vrc-udon` repository at runtime. No hardcoded docs — the remote repo is the single source of truth.
| | |
|---|---|
| **18 MCP tools** | Search, explain, validate, templates, SDK matrix |
| **Dynamic resources** | Skills, rules, cheatsheets, templates |
| **Live indexing** | MiniSearch with weighted ranking + file watcher |
| **IDE support** | Cursor, Claude Desktop, ChatGPT Desktop |
---
## Quick start
**Install into Cursor MCP config** (portable `npx` entry, merge-safe):
```bash
npx -y github:MauDevVR/vrchat-udon-mcp -- install
```
Then **Refresh MCP** in Cursor. Optional: add `--claude` for Claude Desktop.
**Develop from source:**
```bash
git clone https://github.com/MauDevVR/vrchat-udon-mcp.git
cd vrchat-udon-mcp
pnpm install
pnpm update-docs && pnpm build-index && pnpm build
pnpm start
```
See [docs/mcp-config.example.json](docs/mcp-config.example.json) and the language READMEs for workspace, `npx`, global install, submodule, and git dependency options.
---
## Table of contents
- [Documentation](#-documentation)
- [What is this?](#what-is-this)
- [Quick start](#quick-start)
- [Screenshots](#screenshots)
- [Contributing](CONTRIBUTING.md)
- [License](#license)
---
## Screenshots
> Placeholder — add screenshots of MCP tools in Cursor, search results, or validation output here.
| MCP connected in Cursor | `search_documentation` results |
|-------------------------|--------------------------------|
| _Screenshot pending_ | _Screenshot pending_ |
---
## License
[MIT](LICENSE) — Documentation and skills by [niaka3dayo/agent-skills-vrc-udon](https://github.com/niaka3dayo/agent-skills-vrc-udon). MCP server by [MauDevVR](https://github.com/MauDevVR).
TDQS
Scored across 18 tools
Most tools have distinct purposes, but the large number of search tools (e.g., search_documentation, search_reference, search_examples) could cause some confusion despite differentiated descriptions.
The vast majority follow a consistent verb_noun pattern (e.g., explain_topic, list_rules). Only 'sdk_matrix' deviates by omitting a verb, slightly breaking the pattern.
With 18 tools, the set covers a wide range of functionalities for UdonSharp assistance—explanations, validation, templates, rules, and various searches—without feeling excessive or insufficient.
The tool surface appears comprehensive for its domain, covering explainability, validation, templates, rules, skills, SDK info, and multiple search categories with no obvious gaps in read-only knowledge retrieval.