design-review
Skill40 starsChanged 7 months agoMITView on GitHub
Reviews any design document — concept doc, design bible, flow spec, or other architecture-adjacent design artefact — for completeness, internal consistency, and implementability. Detects PRDs and routes them to /prd-review. Use this when the doc is NOT a PRD; for PRD validation, use /prd-review directly.
164 other projects use a version of this.
Adopted by
Most recently taken up by
- Taro000/req-to-designtheir own version, first seen Nov 15, 2026
- persiyanov/skillstheir own version, first seen Nov 15, 2026
- SergioRibera/definitive-uxtheir own version, first seen Nov 15, 2026
- adelrodriguez/skillstheir own version, first seen Nov 15, 2026
- chuanglibs/Claude-Code-Game-Studios-Codextheir own version, first seen Nov 15, 2026
- fergo5002/design-directortheir own version, first seen Nov 15, 2026
Used with
Projects with it also use team-audio, team-level and 3 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
CoralGameSkillteam-audio
Orchestrate audio team: audio-director + sound-designer + technical-artist + gameplay-programmer for full audio pipeline from direction to implementation.
Also in 13 projects that use this
CoralGameSkillteam-level
Orchestrate level design team: level-designer + narrative-director + world-builder + art-director + systems-designer + qa-tester for complete area/level creation.
Also in 12 projects that use this
julianoczkowskiSkillbrief-to-tasks
Break a design brief into an ordered checklist of independently buildable tasks using vertical slices. Saves as a markdown checklist. Use when user wants to break down work, create tasks from a brief, plan implementation order, or mentions "tasks" or "breakdown".
Also in 7 projects that use this
- +2 more from CoralGame
Other versions164 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 164
- garrytan/gstack★ 142ksame name
- Donchitos/Claude-Code-Game-Studios★ 39ksame name
- plugin87/ux-ui-agent-skills★ 3.2ksame name
- DenchHQ/DenchClaw★ 1.7ksame name
- DenchHQ/ironclaw★ 1.7ksame name
- me2resh/apexyard★ 542same name
- julianoczkowski/designer-skills★ 536same name
- openclaw/openclaw-enterprise★ 454same name
- AndrewDryga/emisar★ 376same name
- TheDecipherist/claude-code-mastery-project-starter-kit★ 337same name
154 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository77 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- design/CLAUDE.mdCLAUDE.md
- docs/CLAUDE.mdCLAUDE.md
- src/CLAUDE.mdCLAUDE.md
- tests/CLAUDE.mdCLAUDE.md
- adoptSkill
- architecture-decisionSkill
- architecture-reviewSkill
- asset-auditSkill
- asset-specSkill
- balance-checkSkill
- brainstormSkill
- bug-reportSkill
- bug-triageSkill
- changelogSkill
- code-reviewSkill
- consistency-checkSkill
- content-auditSkill
- create-architectureSkill
- create-control-manifestSkill
- create-epicsSkill
- create-storiesSkill
- day-one-patchSkill
- design-bibleSkill
- design-systemSkill
- dev-storySkill
- estimateSkill
- extractSkill
- gate-checkSkill
- helpSkill
- hotfixSkill
- launch-checklistSkill
- localizeSkill
- map-systemsSkill
- milestone-reviewSkill
- onboardSkill
- patch-notesSkill
- perf-profileSkill
- prd-reviewSkill
- project-stage-detectSkill
- propagate-design-changeSkill
- prototypeSkill
- qa-planSkill
- quick-designSkill
- regression-suiteSkill
- release-checklistSkill
- retrospectiveSkill
- reverse-documentSkill
- review-all-prdsSkill
- scope-checkSkill
- security-auditSkill
- setup-frameworkSkill
- skill-improveSkill
- skill-testSkill
- smoke-checkSkill
- soak-testSkill
- sprint-planSkill
- sprint-statusSkill
- startSkill
- story-doneSkill
- The first 60 shown.
The file
Show its rules and text
We split this skill into the rules it gives an agent
15 rules. 0 are common, 7 are distinctive, and 8 are very specific to this project.
0
Common
Most of the 137 edited versions we read keep these.
7
Distinctive
This author's own choices. Worth a look.
8
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 17
Read-only on the source doc.
- Very specific: line 32
If the user invokes this skill on a design/prd/*.md path, Phase 1 detects that and recommends /prd-review instead — does not duplicate the …
- Distinctive: line 47
Always run in a fresh session — independent of authoring context.
- Very specific: line 74
If path matches design/prd/*.md OR file contains a ## Detailed Requirements section with TR- IDs → this is a PRD.
- Very specific: line 78
If path matches design/ux/*.md OR file is a UX spec → recommend /ux-review <path>.
- Very specific: line 80
If path is design/concept.md → use concept-doc rules (Phase 3a).
- Very specific: line 81
If path is design/design-bible.md → use design-bible rules (Phase 3b).
- Very specific: line 83
If path matches design/flows/*.md → use flow-spec rules (Phase 3c).
- Distinctive: line 140
Every error path resolves to a recoverable state or an explicit unrecoverable terminal.
- Distinctive: line 160
Every entity referenced exists in the registry, OR is declared inside this doc.
- Distinctive: line 162
Every linked file exists.
- Very specific: line 214
Ask: "Write this review to <path>.review.md?"
- Distinctive: line 224
Always runs in a fresh session.
- Distinctive: line 251
Read-only on the source doc.
- Very specific: line 252
Always defers to /prd-review for PRDs and /ux-review for UX specs.
“Common” is measured: kept by at least 6 in 10 of the 137 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Design Review (Generic)
The general-purpose design-document validator. Use this when the artefact
is not a PRD — concept docs, design bibles, flow specs, journey maps,
narrative briefs, etc. For PRDs, route to /prd-review (this skill will
detect that case and tell you).
Read-only on the source doc. Outputs a verdict + blockers list.
Boundary vs /prd-review
| Doc type | Use |
design/prd/*.md | /prd-review (specialised PRD validator) |
design/concept.md | /design-review (concept-doc rules) |
design/design-bible.md | /design-review (visual-system rules) |
design/flows/*.md | /design-review (flow-spec rules) |
design/ux/*.md | /ux-review (defer there) |
Anything else under design/ | /design-review |
If the user invokes this skill on a design/prd/*.md path, Phase 1 detects
that and recommends /prd-review instead — does not duplicate the PRD
validator's work.
Purpose / When to Run
Run when:
- A non-PRD design doc has been authored or revised and needs a sanity pass before downstream work depends on it.
- A concept doc was just produced by
/brainstorm. - A design bible was just produced by
/design-bible. - A flow spec was extracted via
/reverse-document.
Always run in a fresh session — independent of authoring context.
Inputs
- Path to the design doc.
design/concept.mdfor cross-reference.design/registry/entities.yamlfor entity consistency..claude/docs/technical-preferences.md(constraints).- For UI-related docs:
design/design-bible.md.
Outputs
- A printed verdict: APPROVED / APPROVED WITH NOTES / NEEDS REVISION / MAJOR REVISION.
- A blockers + warnings list with section references.
- Optional
<path>.review.mdsnapshot.
Phase 1: Detect Doc Type and Dispatch
Inspect the path and the file's first 30 lines:
- If path matches
design/prd/*.mdOR file contains a## Detailed Requirementssection withTR-IDs → this is a PRD. Stop and recommend/prd-review <path>. Do not duplicate that skill's checks. - If path matches
design/ux/*.mdOR file is a UX spec → recommend/ux-review <path>. Stop. - If path is
design/concept.md→ use concept-doc rules (Phase 3a). - If path is
design/design-bible.md→ use design-bible rules (Phase 3b). - If path matches
design/flows/*.md→ use flow-spec rules (Phase 3c). - Otherwise → use generic design-doc rules (Phase 3d) and warn the user that the rule set is not specialised.
The detection result drives Phase 3.
Phase 2: Read Context
Always read:
- The target doc in full.
design/concept.md(unless the target IS the concept doc).design/registry/entities.yamlfor cross-reference.
For UI-bearing docs additionally read design/design-bible.md.
Phase 3: Run Type-Specific Checks
3a: Concept doc rules
Required sections (per /brainstorm): Premise, Persona, JTBD, Pillars,
Anti-features, Primary Metric, Riskiest Assumption, Constraints.
Quality:
- Premise is one specific sentence (not vague aspiration).
- Persona has a name, a goal, and a current alternative.
- JTBD has when/who/alternative parts.
- Pillars are 3-5 specific design commitments.
- Anti-features list at least 3 things the product will NOT do.
- Primary Metric is measurable (not "engagement" — must be specific).
- Riskiest Assumption is the single thing that, if false, kills the product.
3b: Design-bible rules
Required sections (per /design-bible): Tone, Color, Typography,
Spacing, Iconography, Motion, Imagery, Voice, Component archetypes,
Density, Edge cases, Accessibility, Reference apps.
Quality:
- Color tokens have semantic names AND raw values.
- Typography defines a clear scale with line-heights.
- Motion section gives durations + easing curves.
- Accessibility section addresses contrast, hit targets, motion-reduce, dynamic type, color-blind safety.
3c: Flow-spec rules
Required: trigger, primary path (numbered steps), error paths, success state, abandonment state, analytics events per step.
Quality:
- Each step names the screen and the user action.
- Every error path resolves to a recoverable state or an explicit unrecoverable terminal.
- No "user retries" without specifying retry behavior.
3d: Generic design-doc rules
For unknown doc types:
- Has a clear purpose stated up top.
- Sections are titled and numbered or otherwise navigable.
- Cross-references resolve (linked files exist).
- No
[TBD]/[TODO]left. - Terminology matches the project glossary (CLAUDE.md) and registry.
Warn the user that the rule set is non-specialised and recommend adding a template for this doc type if it's a recurring artefact.
Phase 4: Cross-Reference Checks (all types)
- Every entity referenced exists in the registry, OR is declared inside this doc.
- Every linked file exists.
- The doc does not contradict the concept's anti-features.
- Terminology aligns with the glossary in
app_dev/CLAUDE.md.
Phase 5: Optional Specialist Cross-Check
- Concept docs → optionally spawn
product-directorfor strategic alignment. - Design bibles → optionally spawn
visual-design-director. - Flow specs → optionally spawn
interaction-designer.
The specialist returns findings; the reviewer integrates them.
Phase 6: Verdict and Output
Tally FAIL and WARN:
- 0 FAIL, 0 WARN → APPROVED
- 0 FAIL, 1-5 WARN → APPROVED WITH NOTES
- 0 FAIL, > 5 WARN → NEEDS REVISION
- 1-3 FAIL → NEEDS REVISION
- 4+ FAIL → MAJOR REVISION
Output:
# Design Review: <doc>
**Doc type detected: <concept | design-bible | flow-spec | generic>**
**Verdict: <APPROVED / NEEDS REVISION / MAJOR REVISION>**
## Blockers (FAIL)
1. <issue> — Section <N>
Fix: <action>
## Warnings (WARN)
- <list>
## Strengths
- <list>
## Recommended next steps
- <type-specific guidance>Phase 7: Snapshot
Ask: "Write this review to <path>.review.md?"
Quality Gates
- Verdict matches tally — no soft passes if FAILs exist.
- Phase 1 detection is non-negotiable — if the doc is a PRD, the skill defers and does not run PRD checks itself.
- The reviewer never edits the doc.
- Always runs in a fresh session.
Examples
Example 1 — PRD detection:
/design-review design/prd/auth.md
Phase 1 detects PRD shape. Output: "This file is a PRD. Run
/prd-review design/prd/auth.md instead." Stops.
Example 2 — concept doc:
/design-review design/concept.md
Phase 3a runs. 1 FAIL (Primary Metric is "engagement" — not measurable).
Verdict: NEEDS REVISION. Recommend rerunning /brainstorm to refine
Section 5.
Example 3 — design bible:
/design-review design/design-bible.md
Phase 3b runs. 0 FAIL, 2 WARN (motion durations missing easing curves;
contrast ratio not declared for tertiary text). Verdict: APPROVED WITH
NOTES.
Constraints
- Read-only on the source doc.
- Always defers to
/prd-reviewfor PRDs and/ux-reviewfor UX specs. - The reviewer never silently fixes — surfaces issues for the author.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

