lydakis

lydakis / mcpx

Public

Turn MCP servers into composable CLIs.

14
2
100% credibility
Found Mar 01, 2026 at 13 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Go
AI Summary

mcpx transforms AI tool servers into easy-to-use command-line programs for everyday shell workflows.

How It Works

1
📦 Install mcpx

Download and set up mcpx on your computer with a simple command from your favorite package tool.

2
🔍 Discover your tools

Type mcpx to see a list of ready-to-use AI helpers that work with your existing setups.

3
Add more helpers

Use mcpx add with a link to bring in new tools like GitHub or databases effortlessly.

4
Run your first command

Pick a helper and tool, like mcpx github search-repositories --query=awesome, and get instant results.

5
🔗 Create shortcuts

Install shims so you can type github directly instead of mcpx github every time.

🎉 Mix and match freely

Chain commands with pipes and filters for powerful workflows that feel natural in your terminal.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 13 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is mcpx?

mcpx is a Go CLI that transforms MCP servers—JSON-RPC endpoints for AI tools like GitHub searches or Postgres queries—into Unix-style commands for shell composition. Run `mcpx github search-repositories --query=mcp | jq` to pipe results, or `mcpx ` with native flags from input schemas. Add servers via `mcpx add `, install shims for direct `github search-repositories`, and get bash/zsh completions plus schema-driven `--help`.

Why is it gaining traction?

It bridges MCP's agent-focused APIs to terminal power users, enabling `|`, redirection, and `jq` without custom scripts—unlike verbose MCP SDKs or VS Code plugins. Auto-discovery from Cursor/Claude/Codex setups, optional caching (`--cache=30s`), and shims make it feel native; search lunar mcpx github for multi-turn MCP pipelines or turn FastAPI to MCP endpoints. Devs love the simple contract: list, invoke, compose.

Who should use this?

Terminal-first AI devs building agent workflows, like piping browser-tools or GitHub MCP outputs in scripts. Backend engineers turning GitHub repos into diagrams/prompts/templates via MCP, or those debugging Claude Code servers outside IDEs. Skip if you're GUI-bound or not on MCP yet.

Verdict

Grab it via brew/npm/pip if MCP tools clutter your shell—solid docs and tests make the 10 stars punch above 1.0% credibility weight. Early maturity means test in non-prod; promising for daily driver status. (187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.