My MCP SSE Servers

My MCP SSE Servers is an MCP server orchestration project that manages and runs multiple Model Context Protocol services through Docker Compose. It organizes distinct servers as Git submodules, enabling developers and AI agents to spin up an integrated suite of tools simultaneously. The setup connects to Brave Search for web and local discovery, Playwright for browser automation, Firecrawl for web scraping and crawling, and Grareco for generating graphic recording HTML visuals from text. It also includes the note-mcp service. By wrapping these diverse utilities into a unified container stack exposed over Server-Sent Events (SSE) and streamable HTTP, developers avoid manual process management across separate tool repositories. Engineers and autonomous agent builders use this project to maintain reproducible local testing and production configurations. It standardizes environmental configurations across various agent skills, ensuring consistent tool endpoints are available for LLM clients running automated browser tasks, web data retrieval, and diagram generation workflows.

Category: AI & LLM Tooling

Tags: docker, docker-compose, orchestration, sse

Visit My MCP SSE Servers

How to install and configure My MCP SSE Servers

1. Clone the repository with its submodules: git clone --recurse-submodules https://github.com/iuill/my_mcp_sse_servers 2. Set up the required environment files by copying the sample templates: bash cp brave-search-sse-mcp/.env.sample brave-search-sse-mcp/.env cp mcp_grareco/.env.sample mcp_grareco/.env cp playwright-mcp-docker/.env.sample .env cp firecrawl-sse-mcp/.env.example firecrawl-sse-mcp/.env cp note-mcp-server/.env.example note-mcp-server/.env 3. Start all containerized servers in the background: bash docker-compose up -d 4. In your MCP client (such as Roo-Code), register the servers in your mcpServers configuration using the SSE endpoints: http://host.docker.internal:3101/sse (brave-search-sse), http://host.docker.internal:3102/sse (grareco), http://host.docker.internal:3103/sse (playwright_sse), http://host.docker.internal:3104/sse (firecrawl), and http://host.docker.internal:3105/mcp (note_mcp).

What you can do with My MCP SSE Servers

• Conducting web and local searches using the Brave Search API integration within assistant chat interfaces. * Automating browser interactions and taking screenshots by delegating browser actions to the containerized Playwright service. * Scraping web page contents and crawling site structures through the local Firecrawl SSE container. * Converting arbitrary text or documentation into graphic recording style visual HTML diagrams via Grareco. * Managing persistent notes and documentation workflows through the streamable HTTP note MCP server.

Key facts

• Open Source
• https://github.com/iuill/my_mcp_sse_servers
• AI & LLM Tooling, DevOps, CI/CD & Version Control
• docker, docker-compose, orchestration, sse

Part of MCP Servers

How do I install My MCP SSE Servers?

Clone the repository with its submodules, copy the .env.sample files to .env in each corresponding component directory, and launch the stack using the docker-compose up -d command. Once containers are running, configure your MCP client to access the respective port endpoints.

What can My MCP SSE Servers do?

The project serves multiple agent capabilities concurrently. It provides web search via Brave Search, browser automation via Playwright, web crawling and scraping via Firecrawl, text-to-visual HTML conversion via Grareco, and note interaction services via note-mcp-server.

Which MCP clients work with My MCP SSE Servers?

It works with any MCP client supporting HTTP SSE or streamable HTTP transports, such as Roo-Code, Claude Desktop, and Cursor. You configure the client to communicate directly with the exposed local ports on host.docker.internal.

Is My MCP SSE Servers open source?

Yes, My MCP SSE Servers is an open source project available on GitHub under a community repository structure with submodule dependencies.

How do I stop the running MCP servers?

You can stop all running services by navigating to the project root directory in your terminal and executing docker-compose down, which tears down the containerized network.