Connects streaming data sources like MQTT, Kafka, and Webhooks to AI coding tools via the Model Context Protocol.
Sponsored
mcp-stream is a powerful streaming data engine designed to bridge the gap between real-time data sources and AI coding tools. It ingests data from MQTT, Kafka, and Webhooks, providing a buffered, replayable data pipeline using Redis Streams. Through a comprehensive set of 12 MCP tools, AI agents can manage connections, read live streams, and deploy server-side TypeScript 'Watchers' for tasks like anomaly detection, alerts, and data aggregation, enabling intelligent interaction with dynamic data.
Key Features
01Server-side TypeScript watchers for alerts, anomaly detection, and aggregation
02Redis Streams for buffered and replayable data pipelines
03Multi-source ingestion (MQTT, Kafka, Webhook)
04WebSocket relay with backfill and topic filtering
0512 MCP tools for managing connections, streams, and watchers
062 GitHub stars
Use Cases
01Integrating live sensor data from MQTT into AI coding environments
02Deploying server-side watchers for automated anomaly detection on streaming data
03Connecting Kafka event streams to AI for real-time analysis
