Plan Design Review
Gate AI-generated or agent-built UI with a structured senior-designer scorecard and AI-slop checks before you merge.
Overview
plan-design-review is an agent skill most often used in Ship (also Build, Validate) that scores UI on multiple design dimensions, defines what a 10 looks like, and flags AI Slop before merge.
Install
npx skills add https://github.com/nexu-io/open-design --skill plan-design-reviewWhat is this skill?
- Rates each design dimension on a 0–10 scale with explicit criteria for what a 10 looks like
- Flags AI Slop signals so generic gradient-and-card UIs do not ship unnoticed
- Trigger phrases include plan design review, senior designer review, design rating, and ai slop check
- Curated from Garry Tan gstack creative-direction upstream for Open Design discovery
- Works as a structured checklist alternative to ad-hoc “does this look good?” chat
- 0–10 rating per design dimension
- Explicit AI Slop signal flags
Adoption & trust: 802 installs on skills.sh; 61.4k GitHub stars.
What problem does it solve?
You shipped fast with an agent but cannot tell whether the UI is intentionally designed or interchangeable AI slop worth blocking.
Who is it for?
Solo builders merging agent-generated frontend who want a consistent design gate without hiring a reviewer for every PR.
Skip if: Pure backend or API-only changes, or when a human design lead has already signed off a frozen spec with no visual delta.
When should I use this skill?
User says plan design review, senior designer review, design rating, or ai slop check—or wants a gate before merging UI.
What do I get? / Deliverables
You get dimension-level 0–10 ratings, clear excellence targets, and slop callouts you can fix or reject before merging UI work.
- Per-dimension 0–10 scores
- What-a-10-looks-like notes
- AI Slop flag list
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Canonical shelf is Ship because the skill is framed as a merge gate for UI work—rating dimensions and slop signals right before code lands. Review subphase matches pre-merge design critique rather than initial mock exploration or post-launch analytics.
Where it fits
Score agent-built dashboard chrome and block merge if slop signals dominate.
Iterate on component polish with explicit 10/10 targets per dimension.
Decide whether a landing prototype looks trustworthy enough to run smoke tests.
How it compares
Structured design critique checklist—not a Figma plugin, screenshot tool, or generic code linter.
Common Questions / FAQ
Who is plan-design-review for?
Indie and solo builders using AI coding agents on web or app UI who need a senior-designer-style review rubric before merge.
When should I use plan-design-review?
In Ship review before merging UI; in Build frontend when polishing agent-built screens; in Validate prototype when judging whether a landing or demo looks credible enough to test.
Is plan-design-review safe to install?
Treat it as curated metadata plus upstream gstack; review the Security Audits panel on this Prism page and inspect the upstream repo before installing bundles into your agent skills directory.
SKILL.md
READMESKILL.md - Plan Design Review
# plan-design-review > Curated from Garry Tan (gstack). ## What it does Senior Designer review: rates each design dimension 0-10, explains what a 10 looks like, and flags AI Slop signals. Useful as a gate before merging UI work. ## Source - Upstream: https://github.com/garrytan/gstack - Category: `creative-direction` ## How to use This catalogue entry advertises the skill in Open Design so the agent discovers it during planning. To run the full upstream workflow with its original assets, scripts, and references, install the upstream bundle into your active agent's skills directory: ```bash # Inspect the upstream README for exact paths open https://github.com/garrytan/gstack ``` Then ask the agent to invoke this skill by name (`plan-design-review`) or with one of the trigger phrases listed in this skill's frontmatter.