跳到主要内容

MCP Server

The BlendVision MCP Server is a Model Context Protocol (MCP) server that enables AI assistants — such as Claude, Cursor, and VS Code Copilot — to interact directly with BlendVision One APIs. With 35+ built-in tools spanning VOD, live streaming, chatrooms, analytics, and more, you can manage your video platform through natural language conversations.

Prerequisites​

  • Node.js 18 or later
  • A valid BlendVision API token — see Authentication for how to obtain one

Installation​

Using npx (no installation required)​

npx @blendvision/mcp-server

Global installation via npm​

npm install -g @blendvision/mcp-server

Build from source​

git clone https://github.com/BlendVision/mcp-server.git
cd mcp-server
npm install
npm run build

Configuration​

Environment variables​

VariableRequiredDescription
BLENDVISION_API_TOKENYesYour BlendVision API token
BLENDVISION_BASE_URLNoAPI base URL (defaults to https://api.one.blendvision.com)

Claude Desktop​

Add the following to your Claude Desktop configuration file.

  • macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
  • Windows: %APPDATA%\Claude\claude_desktop_config.json
{
  "mcpServers": {
    "blendvision": {
      "command": "npx",
      "args": ["-y", "@blendvision/mcp-server"],
      "env": {
        "BLENDVISION_API_TOKEN": "<your-api-token>"
      }
    }
  }
}

Claude Code​

In the project root, create or edit .mcp.json:

{
  "mcpServers": {
    "blendvision": {
      "command": "npx",
      "args": ["-y", "@blendvision/mcp-server"],
      "env": {
        "BLENDVISION_API_TOKEN": "<your-api-token>"
      }
    }
  }
}

Cursor / VS Code​

Add the server under the MCP settings in your IDE configuration. Refer to the MCP protocol documentation for IDE-specific setup guides.

Available tools​

Video On Demand​

ToolDescription
list_videosList all videos with filtering options
get_videoGet video details by ID
create_videoCreate a new video
update_videoUpdate video metadata
delete_videoDelete a video
update_vod_subtitlesUpdate subtitles for a VOD video

Live Streaming​

ToolDescription
list_live_channelsList all live channels
get_live_channelGet channel details by ID
create_live_channelCreate a new live channel
update_live_channelUpdate channel settings
delete_live_channelDelete a channel
start_liveStart live streaming on a channel
stop_liveStop live streaming on a channel
cancel_live_channelCancel and archive a live stream
archive_live_channelArchive a live stream permanently

Library & File Upload​

ToolDescription
upload_fileInitiate file upload and get presigned URLs
complete_file_uploadComplete a file upload session after uploading parts

Chatroom​

ToolDescription
list_chatroomsList all chatrooms
get_chatroomGet chatroom details
create_chatroomCreate a new chatroom
send_chatroom_messageSend a message to a chatroom

Account & Organization​

ToolDescription
get_accountGet current account information
list_organizationsList accessible organizations

Playback​

ToolDescription
generate_playback_tokenGenerate a playback token for content access
list_playback_codesList playback codes for a resource

Analytics​

ToolDescription
get_analyticsRetrieve analytics reports with various report types

Example prompts​

Once connected, you can interact with BlendVision through natural language:

  • "List all my videos"
  • "Create a new live channel called 'Product Launch'"
  • "Generate a playback token for video ID abc123"
  • "Show me analytics for the last 7 days"
  • "Upload a video file and set its title to 'Demo'"
  • "Send a welcome message to the main chatroom"

Architecture​

The MCP server follows a modular, layered design:

  • Protocol Layer — handles MCP protocol routing and message parsing
  • Tool Registry — centralized tool registration and discovery
  • Tool Modules — domain-specific implementations (VOD, Live, Chatroom, etc.)
  • Base Tool Layer — shared features such as retry logic, pagination, and error handling
  • Client Layer — HTTP communication with the BlendVision One API

Testing​

MCP Inspector​

Use the built-in inspector to test tools interactively via a web interface:

npm run inspector

Connection test​

Verify that your API credentials are valid:

export BLENDVISION_API_TOKEN="<your-api-token>"
npm run test:connection

Resources​