seargo

Seargo is a minimal metasearch engine designed for personal use. It exposes a search endpoint and an MCP (Meta Collaboration Protocol) interface to integrate with AI agents. The server runs inside a Docker container and serves a web-based search page at /search?q=QUERY, as well as a streaming MCP endpoint at /mcp and a Server-Sent Events (SSE) endpoint at /sse for live results. You can use the MCP endpoints to query multiple providers ( Bing, Brave, DuckDuckGo, Startpage, and English Wikipedia as of now ) and have AI agents co-ordinate across sources. To start using it, run the provided Docker command, then access http://127.0.0.1:8080/search?q=YOUR_QUERY for human-friendly search results or connect an MCP-enabled agent to http://127.0.0.1:8080/mcp for programmatic queries. The SSE endpoint at http://127.0.0.1:8080/sse can be used for real-time streaming of results to clients that support it. The server aggregates results from the configured providers and returns a unified set of hits.

Prerequisites:

• Docker installed on your machine
• Internet access to pull the container image

Installation steps:

1. Ensure Docker is running on your system.

2. Run the container using the recommended command:

   docker run -d --rm -p 8080:8080 ghcr.io/netroy/seargo

3. Wait a moment for the container to start. Verify it’s listening on port 8080:

   curl -s http://127.0.0.1:8080/search?q=test

4. Optional: Update firewall rules or listen address if you’re running on a server. You may also map a different host port if 8080 is in use, e.g. -p 8081:8080 and access http://127.0.0.1:8081/search?q=.

5. If you need to stop the container later:

   docker ps | grep seargo docker stop <container_id>

Notes:

• The container exposes /search for human users and /mcp for MCP-enabled agents. /sse provides a streaming interface for real-time results.
• Providers currently include Bing, Brave, DuckDuckGo, Startpage, and English Wikipedia. If you need more sources, you may need to adjust the provider configuration inside the Seargo deployment.
• If you run into port conflicts, re-map ports with -p hostPort:8080 in the docker run command.
• Ensure your environment allows outbound HTTP requests to the provider endpoints; some networks may require proxy configuration.