Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Agent skills to help you build products your users love.
| Date | Stars |
|---|---|
| 2026-07-31 | 487 |
| 2026-08-05 | 490 |
| 2026-08-06 | 490 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# 🦆 Produck Skills **High-quality, user-experience-centered skills for your coding agents.** This repository is our open-source, documentation-driven collection of [agent skills](https://code.claude.com/docs/en/skills) to help you build products users love. Each skill is a folder of markdown that teaches a coding agent (Claude Code, Cursor, Codex, Copilot, and others) how to ship features and fixes that your users want to improve product-market fit — starting with turning messy user requests into context-rich PRDs your agents can actually execute. <picture> <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/chart?repos=tryproduck/produck-skills&type=date&theme=dark&legend=top-left" /> <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/chart?repos=tryproduck/produck-skills&type=date&legend=top-left" /> <img alt="Star History Chart" src="https://api.star-history.com/chart?repos=tryproduck/produck-skills&type=date&legend=top-left" /> </picture> > ⭐ **If these skills are useful, [star the repo](https://github.com/tryproduck/produck-skills)** — it helps other builders find it. --- ## Install One line installs every skill in this repo into your coding agent: ```bash npx skills add tryproduck/produck-skills ``` That's it. It uses the open [`skills`](https://github.com/vercel-labs/skills) CLI (no global install, just `npx`) and works across 70+ agents — Claude Code, Cursor, Codex, Copilot, and more. The skills are available the next time your agent starts. --- ## Skills Catalog | Skill | What it does | | --- | --- | | [`user-alignment`](skills/user-alignment/SKILL.md) | Turn vague or messy user requests into aligned, agent-executable PRDs — with scope, bounded phases, testable acceptance criteria, and explicit do-not-do boundaries. | | [`produck-feedback`](skills/produck-feedback/SKILL.md) | Act on real in-context user feedback: pull a feedback ticket's full context through the Produck MCP and drive a feedback → aligned PRD → build loop. Pairs with `user-alignment`. | | [`demo-studio`](skills/demo-studio/SKILL.md) | Build a premium, fully-animated product-demo / launch video from a written brief. Scene-by-scene motion graphics with camera moves, real UI mockups, and a complete original sound kit, rendered from HTML + CSS + GSAP to MP4 via HyperFrames. Ships a runnable example film and a 25-sound synthesized SFX kit. | | [`taste`](skills/taste/SKILL.md) | Improve website visual taste with a screenshot-driven hierarchy, trust, spacing, CTA, and consistency loop — so agents ship visibly better UI instead of generic polish. | More user-experience skills are on the way. --- ## Taste benchmark We ran a one-shot website improvement benchmark on the same intentionally flawed fictional SaaS landing page: 1. **Baseline** — feature-led platform copy, generic dashboard UI, weak proof, and noisy hierarchy. 2. **Taste-guided improvement** — the same page rebuilt with the `taste` skill's category thesis, worldview sequence, domain artifacts, rubric, and screenshot verification loop. Screenshots were captured at `1440 × 1100` with `npx -y chrome-devtools-axi` and combined into one before/after asset so the improvement is reviewable at a glance.  | Variant | Category idea | Worldview | Hierarchy | Clarity | Trust | Rhythm | CTA gravity | Domain specificity | Consistency | | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: | | Baseline | 1 | 1 | 1 | 2 | 1 | 1 | 2 | 1 | 1 | | Taste-guided improvement | 5 | 5 | 5 | 5 | 4 | 5 | 5 | 5 | 5 | The taste-guided pass is built around one category idea — **the judgment layer for product feedback** — and opens with the worldview that feedback is abundant while judgment is scarce. It replaces fake dashboard polish with domain artifacts: opportunity memos, accepted/rejected evidence, and decision briefs.
Excerpt of 6,614 characters
Read on GitHubWould you bet a product on this? Bounded 0–100 and slow moving.
matched fp:d05a7e2bfea1d213, topic:developer-tools