luml-ai

luml-ai / AGENTS.lock

Public

A package manager for Agents/Skills/MCPs

20
0
100% credibility
Found Feb 04, 2026 at 18 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

AGENTS.lock is a tool for managing and synchronizing AI agent skills, configurations, and instructions across multiple AI command-line clients using a single lockfile.

How It Works

1
💡 Hear about the helper

You learn about a simple tool that keeps all your AI assistants sharing the same skills and instructions across different programs.

2
🛠️ Get it ready

You add the tool to your computer with one easy step, so it's there whenever you need it.

3
📝 Start your list

You create a single list file in your project that notes down all the AI skills, helpers, and guidance you want to use.

4
Pick your favorites

You easily add links to cool AI skills, smart agents, or helpful notes, and the tool remembers them all in one place.

5
🔄 Share with everyone

You run a quick update, and the tool copies your list to make sure all your AI programs have the same setup.

🎉 Everything in sync

Now your AI assistants across different tools all know the same skills and instructions, making your work smoother and faster.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 20 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 AGENTS.lock?

AGENTS.lock is a Python package manager for AI agent configs, using a TOML lockfile as the single source of truth for skills, agents, MCP servers, and instructions across Claude, Codex, Gemini, and Copilot CLI tools. Install via pip, run `al init` to create AGENTS.lock, `al add skill https://github.com/...` to pull from repos, and `al sync` to deploy to client directories. It solves the mess of manually syncing configs between fragmented AI CLIs, like a package manager python tailored for agents.

Why is it gaining traction?

It unifies management like npm's package.json or github package manager but for AI skills, with git pinning, project/global scopes, and aliases for multi-lockfile workflows. `al sync` auto-renders client-specific formats, handles updates, and prunes orphans—no more copy-pasting or version drift. Devs dig the console-first flow on Linux, macOS, Windows, bridging github package registry python gaps for agent storage.

Who should use this?

AI workflow builders using Claude/Copilot in teams, needing shared skills/MCPs without per-tool tweaks. Solo devs managing agent libs via GitHub URLs, or package manager console users on Ubuntu/Windows tired of siloed configs. Ideal for lock-free agents in fast gametime prototyping.

Verdict

Early alpha (18 stars, 1.0% credibility) with breaking changes warned—test coverage and docs are basic but CLI shines. Grab it if multi-AI CLIs dominate your stack; skip for production until stabilized.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.