dhanababum

dhanababum / mcpskills-cli

Public

Generate Agent Skills from MCP server tools. Connects via Streamable HTTP, discovers tools, and outputs a skill with schema docs and a call script in the language of your choice.

adk-google agents chatgpt claude-code cline
14
1
100% credibility
Found Feb 18, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

This project is a tool for creating AI agent skills with documentation and call scripts from tools on an MCP server.

How It Works

1
🔍 Discover the skill maker

You find a helpful tool that turns actions from your server into easy skill packs for AI helpers, saving space and making them smarter.

2
📥 Set up the tool

You add this simple program to your computer so it's ready to create skills anytime.

3
🔗 Link your server

Share your server's location and a private password so it can grab the list of helpful actions available.

4
Choose skill style
📦
One big skill

Combine everything into a single easy folder with all actions documented.

📂
Separate skills

Make individual folders, one for each action, for precise control.

5
✨ Create the skills

The tool builds folders with clear guides explaining each action and ready-to-use shortcut scripts.

✅ Skills ready!

Your AI helper now accesses these skills on demand, working faster without clutter or extra cost.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 12 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 mcpskills-cli?

This Python CLI tool connects to MCP servers over Streamable HTTP, discovers available tools, and generates static agent skills—complete with SKILL.md docs outlining schemas and executable call scripts in bash, Python, Node, Go, or Rust. It solves the token bloat problem of loading full MCP toolsets into agent context every time, instead creating on-demand skills that slash usage (demo shows 50-90% savings). Run `mcpskills-cli --url --token --script python` to output ready-to-use files in `~/.cursor/skills` for agent generate instructions or copilot setups.

Why is it gaining traction?

Unlike raw MCP polling that wastes tokens on unused tools, it bakes skills into lightweight markdown plus scripts, making agents leaner and editable without regenerating. The `--multi-skills` flag splits tools into separate dirs for granular loading, and auto-saved credentials mean easy token rotation. Developers dig the one-command workflow to generate agents.md prompts from any MCP endpoint.

Who should use this?

AI agent builders exposing tools via MCP servers, like backend devs running local DB or graph APIs for Cursor agents. Perfect for teams generating agent instructions copilot-style or prototyping agent leads reviews without token penalties. Skip if you're not in the MCP ecosystem.

Verdict

Grab it for MCP-to-skills workflows—solid docs and CLI make it dead simple despite 12 stars and 1.0% credibility score signaling early days. Niche maturity means test your MCP server first, but token wins make it worth the pip install.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.