@streamflux/mcp-server

Connect AI agents to real-time data streams via MCP (Model Context Protocol)

Let your AI assistant monitor, analyze, and interact with real-time data.

Features

Core Tools

• list_topics - List all active topics in your StreamFlux server
• get_stats - Get server statistics (connections, throughput, uptime)
• publish_event - Publish events to any topic
• subscribe_sample - Sample messages from a topic for analysis
• create_window - Create time-based windows for event aggregation
• health_check - Check server connectivity

AI-Native Tools

• analyze_stream - Analyze topic patterns, message frequency, and health status
• detect_anomalies - Monitor topics for unusual activity (spikes, drops)
• broadcast_ai_event - Publish AI-tagged events with confidence scores
• get_stream_summary - Get AI-friendly overview of all streams with suggested actions

Installation

npm install -g @streamflux/mcp-server
# or
pnpm add -g @streamflux/mcp-server

Usage with Cursor

Add to your Cursor MCP settings (~/.cursor/mcp.json):

{
  "mcpServers": {
    "streamflux": {
      "command": "npx",
      "args": ["-y", "@streamflux/mcp-server"],
      "env": {
        "STREAMFLUX_URL": "https://api.streamflux.dev",
        "STREAMFLUX_API_KEY": "your-api-key"
      }
    }
  }
}

Or with local development server:

{
  "mcpServers": {
    "streamflux": {
      "command": "npx",
      "args": ["-y", "@streamflux/mcp-server"],
      "env": {
        "STREAMFLUX_URL": "http://localhost:8080",
        "STREAMFLUX_API_KEY": "your-api-key"
      }
    }
  }
}

Usage with Claude Desktop

Add to your Claude Desktop config:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "streamflux": {
      "command": "npx",
      "args": ["-y", "@streamflux/mcp-server"],
      "env": {
        "STREAMFLUX_URL": "https://api.streamflux.dev",
        "STREAMFLUX_API_KEY": "your-api-key"
      }
    }
  }
}

Environment Variables

| Variable | Description | Default | |----------|-------------|---------| | STREAMFLUX_URL | StreamFlux server URL | https://api.streamflux.dev | | STREAMFLUX_API_KEY | API key for authentication | (optional) |

Example Prompts

Once configured, you can ask your AI assistant:

Basic Operations

"List all active topics in StreamFlux"

"Publish a test event to the 'orders' topic"

"Show me the server statistics"

AI-Native Operations

"Analyze the orders stream - is it healthy?"

"Detect any anomalies in my data streams"

"Broadcast an alert saying 'High CPU detected' to the notifications topic"

"Give me a summary of all streams with recommended actions"

Resources

The MCP server also exposes topics as resources. You can ask:

"What resources are available from StreamFlux?"

Each topic is exposed as a resource with URI format: streamflux://topics/{topic-name}

Development

# Install dependencies
pnpm install

# Run in development mode
pnpm dev

# Build for production
pnpm build

License

MIT