Leeroo-AI

An ML engineering plugin for your coding agents.

18
3
100% credibility
Found Mar 13, 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

SuperML is a plugin that equips AI coding assistants with machine learning workflow skills and access to a vast knowledge base for planning, verifying, debugging, and optimizing ML projects.

How It Works

1
🔍 Discover SuperML

You hear about SuperML while looking for ways to make your AI coding helper smarter at machine learning tasks, and watch a quick demo video showing it in action.

2
🛠️ Pick Your AI Helper

You choose your favorite AI coding tool like Claude or Cursor where you want to add these ML superpowers.

3
Add the Helper
🔑
Connect Extra Smarts

Optionally sign up for a free trial key to unlock a huge library of ML know-how for even better results.

🚀
Jump Right In

Start using it right away with built-in web searches if you skip the key.

4
💬 Chat About ML Work

You start a conversation with your AI helper about your machine learning project, like planning a training setup or fixing an error.

5
🧠 Get Expert Advice

Your AI now plans step-by-step, checks for mistakes, debugs issues, and suggests improvements using proven ML knowledge, saving you hours of trial and error.

6
📈 Build and Improve

You follow the advice to set up your project, run tests, and iterate as your AI remembers what worked before across chats.

🎉 ML Project Succeeds

Your machine learning model trains perfectly without crashes or wasted time, and you feel like you have a pro ML engineer on your team.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 18 to 18 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 superml?

SuperML is a Python plugin that supercharges AI coding agents like Claude Code, Cursor, and OpenCode with ML engineering workflows. It injects seven skills for planning pipelines, verifying configs, debugging issues like OOM or NaN, and iterating experiments, plus a persistent memory backed by a 27k-page knowledge base on frameworks from vLLM to DeepSpeed. Users get grounded configs with doc citations, experiment tracking across sessions, and an ml-expert agent for deep dives—no more hallucinated hyperparameters wasting GPU time.

Why is it gaining traction?

Unlike generic agents that guess ML setups, SuperML enforces real doc-grounded steps, boosting judged performance from 8.3/15 to 13.2/15 on tasks like QLoRA and DPO, with 91% win rate. The hook is instant value: install via `/plugin install superml@leeroo-marketplace`, prompt your agent for a training config, and it catches pitfalls before launch. It stands out among engineering plugins like noise engineering plugins or compounding engineering plugin github tools by compounding knowledge via Leeroopedia lookups.

Who should use this?

ML engineers fine-tuning LLMs on A100s, optimizing inference with quantization, or building RAG/DPO pipelines in agents. Ideal for solo devs handling distributed training or debugging divergence, who use Claude Code for compound engineering plugin claude code workflows but hate repeating failed experiments.

Verdict

Worth a quick install for ML-heavy agent users—docs are solid, tests cover real scenarios—but with 18 stars and 1.0% credibility score, it's early-stage; expect occasional rough edges. Try the free API credits first.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.