ML-SystemDesign · @ML-SystemDesign ↗ Machine Learning System Design
Agent skills that bring the ML System Design framework to your AI coding agent: grade a design doc or repo, then decide whether an AI product has earned its next gate.
The roster
Review ML System Design
ML-SystemDesign Point your agent at an ML design doc, a repo, or both and get a graded scorecard: a verdict, per-dimension grades, severity-ranked findings, and the cheapest high-impact fixes first.
MLSystemDesign/ml-system-design-review 775 
Review AI Product Idea
ML-SystemDesign Run a stage-gate review on any AI product and get a Go, Conditional, or Kill decision backed by a scored readiness table, the top blocker, and the next-stage investment to release.
MLSystemDesign/ai-stage-gate 775 What it does
Point your AI coding agent at an ML or AI system design and get the framework from Kravchenko and Babushkin's book applied for you, not a generic checklist.
One skill reviews a design doc, a repo, or both and returns a verdict, a stage-aware gradecard, severity-ranked findings, and a prioritized fix plan. Another runs a stage-gate review and reaches an evidence-based Go, Conditional, or Kill decision for an AI product. The repo also ships the book's design-doc templates, a review checklist, and worked example documents.
You get concrete, shareable output you can paste into a design doc or a gate deck, grounded in what your design or repo actually shows.
When to use it
- You're reviewing an ML or AI system design
Grade a design doc, a repo, or both against the book's framework and get findings, a gradecard, and the cheap fixes worth doing first.
- You need a Go or Kill call at an AI product gate
Run a stage-gate review and get an evidence-based decision with the top blocker and the exact next-stage investment to release.
- You're applying the book to a real project
Use the same rubrics, templates, and worked examples the book teaches, run by your agent against your own design instead of on paper.
Common questions
What's included in this collection?
Machine Learning System Design ships two agent skills, one for design review and one for AI stage-gate decisions, plus the book's design-doc templates, a review checklist, and worked example design documents.
Do I install each skill separately?
No. One install, npx skills add ML-SystemDesign/MLSystemDesign, brings the whole set. You can then invoke any skill on its own without the rest.
Do I need the book to use these skills?
No. The skills stand alone and work without the book. They apply the ML System Design framework by Kravchenko and Babushkin, so the Manning book goes deeper on the same concepts if you want it.
Is it free and open source?
Yes. Machine Learning System Design is MIT-licensed and free to use. You run the skills locally in your own agent, against your own designs.