Show HN: MCP-stama – An ultra-fast Rust MCP server with no dependencies

The Ultra-Fast Swiss Army Knife MCP Server for AI Coding Agents

Blazingly fast, zero-dependency, single static Rust binary built for Cursor, Claude Desktop, and Windsurf.

⚡

Why switch toStandard Node.js or Python MCP tools consumemcp-stama

?200MB+ RAMand take1–3 secondsjust to wake up.mcp-stama

runs natively in under10MB RAMwithsub-millisecond tool response times, keeping your AI agent fast and your computer cool.

Each tool in mcp-stama

is engineered to give your AI agent deep local context without wasting prompt tokens or spawning heavy child processes:

• 🔎 : Sub-millisecond file search and regex scanning usingfast_grep

ignore::WalkBuilder

. Respects.gitignore

and.ignore

automatically while skipping binary/hidden files. - 📦 : Pure-Rust Git inspector powered natively bygit_snapshot

gix

(gitoxide

). Fetches HEAD commits, branch information, line deltas, and file status without calling the externalgit

executable. - 🐳 : Instant host and container diagnostics powered bydocker_watcher

bollard

andsysinfo

. Gives your AI instant visibility into running Docker containers, system memory, CPU cores, and mapped ports. - ⚡ : Includesauto-configurator

--install-cursor

and--install-claude

flags to automatically injectmcp-stama

into your editor settings in under 2 seconds. - 🛡️ Pure Stdio JSON-RPC 2.0: All diagnostic and logging outputs are routed strictly tostderr

, keepingstdout

100% compliant for JSON-RPC frame protocol traffic.

curl -fsSL [https://raw.githubusercontent.com/StamManif/mcp-stama/main/install.sh](https://raw.githubusercontent.com/StamManif/mcp-stama/main/install.sh) | sh
iwr -useb [https://raw.githubusercontent.com/StamManif/mcp-stama/main/install.ps1](https://raw.githubusercontent.com/StamManif/mcp-stama/main/install.ps1) | iex
cargo install mcp-stama

Skip manual JSON editing! Let mcp-stama

configure your AI client automatically:

MacOS/Linux

{ command -v mcp-stama >/dev/null 2>&1 && mcp-stama || ~/.local/bin/mcp-stama; } --install-claude

Windows

& "$HOME\.mcp-stama\bin\mcp-stama.exe" --install-claude

MacOS/Linux

{ command -v mcp-stama >/dev/null 2>&1 && mcp-stama || ~/.local/bin/mcp-stama; } --install-cursor

Windows

& "$HOME\.mcp-stama\bin\mcp-stama.exe" --install-cursor

If you prefer adding it manually to your settings file:

{
  "mcpServers": {
    "mcp-stama": {
      "command": "mcp-stama",
      "args": []
    }
  }
}
{
  "mcpServers": {
    "mcp-stama": {
      "command": "mcp-stama",
      "args": []
    }
  }
}

Run the embedded benchmark suite on your own machine at any time with mcp-stama --benchmark

:

                 +-------------------------------+
                 |  AI Client (Cursor / Claude)  |
                 +---------------+---------------+
                                 |
                          stdio (JSON-RPC)
                                 |
                 +---------------v---------------+
                 |        StdioTransport         |
                 +---------------+---------------+
                                 |
                     JsonRpcRequest / Response
                                 |
                 +---------------v---------------+
                 |          ToolRegistry         |
                 +---------------+---------------+
                                 |
         +-----------------------+-----------------------+
         |                       |                       |
+--------v--------+     +--------v--------+     +--------v--------+
|    fast_grep    |     |  git_snapshot   |     | docker_watcher  |
| (ignore/regex)  |     |   (gix/rust)    |     | (bollard/sys)   |
+-----------------+     +-----------------+     +-----------------+
git clone [https://github.com/StamManif/mcp-stama.git](https://github.com/StamManif/mcp-stama.git)
cd mcp-stama

cargo test

cargo build --release

./target/release/mcp-stama --benchmark

Licensed under the Apache License, Version 2.0 (LICENSE or http://www.apache.org/licenses/LICENSE-2.0).