1. Conduid
  2. AI
  3. vrchat
MCP server · AI

vrchat

VRChat MCP Server - Access VRChat API through Model Context Protocol

Unclaimed ai
39Low

Scored 5 months ago · breakdown

About vrchat

vrchat is an MCP server published by git+sawa-zen in the AI category: vRChat MCP Server - Access VRChat API through Model Context Protocol.

Install

Install
npx vrchat-mcp
Claude Code
claude mcp add vrchat-mcp-gitsawaz -- npx -y vrchat-mcp
npx
npx -y vrchat-mcp

This server has no ConduID identity, so agent calls to it are not receipted. Pin the version you install and review the source before granting it credentials.

Ask AI

✦

Ask AI about vrchat

Powered by Claude · Grounded in docs

I know everything about vrchat. Ask me about installation, configuration, usage, or troubleshooting.

Security checks

  • ✓README presentThe repository ships documentation.
  • !Scoped permissionsDoesn't declare a permission scope. Assume it can do anything its process can.

4 checks have not been run yet.

README

VRChat MCP

npm version

日本語 | 한국어

This project is a Model Context Protocol (MCP) server for interacting with the VRChat API. It allows you to retrieve various information from VRChat using a standardized protocol.

Overview

The VRChat MCP server provides a way to access VRChat's API endpoints in a structured manner. It supports a wide range of functionalities, including user authentication, retrieving user and friend information, accessing avatar and world data, and more.

Usage

To start the server, ensure you have the necessary environment variables set:

export VRCHAT_USERNAME=your_username
export VRCHAT_AUTH_TOKEN=your_auth_token

[!NOTE]

How to obtain AUTH TOKEN

You can use the following command to login and obtain an auth token:

$ npx vrchat-auth-token-checker

VRChat Username: your-username
Password: ********

# If 2FA is enabled
2FA Code: 123456

# Success output
Auth Token: authcookie-xxxxx

Command source code

Please handle the obtained token with care as it has a very long lifetime

Then, run the following command:

npx vrchat-mcp

This will launch the MCP server, allowing you to interact with the VRChat API through the defined tools.

Usage with Claude Desktop

To use this MCP server with Claude Desktop, you do not need to run npx vrchat-mcp manually. Instead, add the following configuration to your Claude Desktop config file:

  • MacOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "vrchat-mcp": {
      "command": "npx",
      "args": ["vrchat-mcp"],
      "env": {
        "VRCHAT_USERNAME": "your-username",
        "VRCHAT_AUTH_TOKEN": "your-auth-token"
      }
    }
  }
}

Then, start Claude Desktop as usual. If you have to use nodenv or nvm, you may need to specify the full path to the npx command.

Available Tools

This Model Context Protocol server provides the following VRChat-related tools:

User Related

  • vrchat_get_friends_list: Get a list of friends
  • vrchat_send_friend_request: Send a friend request

Avatar Related

  • vrchat_search_avatars: Search for avatars
  • vrchat_select_avatar: Select and switch to a specific avatar

World Related

  • vrchat_search_worlds: Search for worlds
  • vrchat_list_favorited_worlds: Get a list of favorited worlds

Instance Related

  • vrchat_create_instance: Create a new instance
  • vrchat_get_instance: Get information about a specific instance

Group Related

  • vrchat_search_groups: Search for groups
  • vrchat_join_group: Join a group

Favorites Related

  • vrchat_list_favorites: Get a list of favorites
  • vrchat_add_favorite: Add a new favorite
  • vrchat_list_favorite_groups: Get a list of favorite groups

Invite Related

  • vrchat_list_invite_messages: Get a list of invite messages
  • vrchat_request_invite: Request an invite
  • vrchat_get_invite_message: Get a specific invite message

Notification Related

  • vrchat_get_notifications: Get a list of notifications

Debugging

First, build the project:

npm install
npm run build

Since MCP servers run over stdio, debugging can be challenging. For the best debugging experience, we strongly recommend using the MCP Inspector.

You can launch the MCP Inspector via npm with this command:

npx @modelcontextprotocol/inspector "./dist/main.js"

Be sure that environment variables are properly configured.

Upon launching, the Inspector will display a URL that you can access in your browser to begin debugging.

Publishing

To publish a new version of the package, follow these steps:

  1. Pull the latest code from the main branch

    git checkout main
    git pull origin main
    
  2. Build the package

    npm run build
    
  3. Publish to npm

    npm publish
    
  4. Push changes to the remote repository

    git push origin main --tags
    

Contributing

Contributions are welcome! Please fork the repository and submit a pull request for any improvements or bug fixes.

License

This project is licensed under the MIT License. See the LICENSE file for details.

README copy captured when this server was indexed; it may lag the repository. The original is authoritative.

Questions

About vrchat

How do I install vrchat?

Run npx vrchat-mcp, then add the server to your MCP client's configuration.

Is vrchat safe to use with an AI agent?

Its trust score is 39 out of 100 (low). It passes 1 of the 2 static security checks run so far; the failures are listed above. It has no ConduID identity yet, so agent calls to it are not receipted.

Is vrchat still maintained?

Conduid hasn't recorded a commit date for this repository yet. Check the repository directly for recent activity.