theo-feng03/ai_pm-skills — explained in plain English
Analysis updated 2026-05-18
Walk through a complex AI product idea module by module to produce structured planning material.
Convert an existing product idea or competitor research into a coding-agent-ready Markdown requirements document.
Prepare a clear MVP scope and acceptance criteria before handing a project to a coding agent.
| theo-feng03/ai_pm-skills | 0pen-sourcer/hearth | 0rnot/god-mode-claude | |
|---|---|---|---|
| Stars | 15 | 15 | 15 |
| Language | — | Python | — |
| Setup difficulty | easy | moderate | easy |
| Complexity | 2/5 | 3/5 | 1/5 |
| Audience | pm founder | vibe coder | developer |
Figures from each repo's GitHub metadata at analysis time.
No license has been specified yet, so reuse terms are undefined.
AI PM Skills is a small collection of two reusable Skills meant for AI product managers and AI product builders. The idea is to help turn a vague AI product idea into something structured enough to discuss, plan, and hand off to a coding agent like Claude Code, Codex, or similar tools. The first Skill, ai-prd-copilot, works like a product engineering partner that walks through a complex AI product idea module by module instead of writing a full requirements document all at once. It asks about target users and scenarios, business goals, technical architecture, page design, task states, human review steps, prompt and policy management, feedback handling, permissions and compliance, cost budgeting, and staged rollout. Each module produces a chunk of raw material that can later be assembled into a formal product requirements document, written from a perspective that considers system boundaries and failure handling, not just what a user wants. The second Skill, vibe-coding-prd, has a different goal: converting an idea, an existing requirements document, or competitor research into a Markdown document that a coding agent can act on directly. It focuses on defining MVP scope, feature modules, page structure, a recommended tech stack, data objects and states, and concrete acceptance criteria for each feature, along with an optional starting prompt for the coding agent. The two Skills are meant to be used together or separately, and the README describes several example workflows, from taking a rough idea through both Skills in sequence, to using just one Skill on an already clear idea. Each Skill is organized as a set of markdown reference files plus a small configuration folder. The README plainly states that no license has been specified yet, so the terms for reusing this project are not defined.
Two reusable AI product manager Skills that turn a vague AI product idea into structured planning material and coding-agent-ready requirements documents.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly pm founder.
This repo across BitVibe Labs
Verify against the repo before relying on details.