jubscodes/get-shit-pretty/gsp-doctor
Claude Code and agent skill by jubscodes/get-shit-pretty
- community
Checking that the SKILL.md file is still in the repository.
Check project health — use when: something's broken, check the project, is everything set up right, health check, what's the status of my GSP setup
Install
npx vskill@latest install jubscodes/get-shit-pretty/gsp-doctorSKILL.md
<context> Diagnostic tool for GSP design projects. Pure pattern matching — no agents, no file writes. </context>
<objective> Health check on .design/ — all brands and projects. Terminal-only output.
Checks: structure, phase ordering, stale outputs, config drift, missing chunks, broken references, review status, brand drift, upgrade detection </objective>
<process>
Step 0: Find design directory
Check for .design/ in the current directory.
If not found:
Stop here.
Step 1: Detect structure type
New dual-diamond structure: .design/branding/ or .design/projects/ exists Legacy flat structure: .design/config.json exists at root (not inside branding/ or projects/) Empty: .design/ exists but has neither
For legacy: run legacy checks (same as v0.3.0 doctor). For new: run multi-instance checks below.
Step 2: Scan all instances
Brands: List all directories in .design/branding/ that have a config.json with project_type: "brand" Projects: List all directories in .design/projects/ that have a config.json with project_type: "design"
For each instance, read:
config.json— configurationSTATE.md— phase progressBRIEF.md— briefbrand.ref— brand reference (projects only)
Step 3: Run checks per instance
Load ${CLAUDE_SKILL_DIR}/checks.md — the full check catalog with pass/warn/fail criteria. Read it once at this step; do not load it before Step 2 because the applicable check sets depend on what was scanned.
Run only the sets applicable to the detected state:
| Detected | Run | |
|---|---|---|
| Each brand instance | B1, B2, B3, B4 | |
| Each project instance | P1–P10 | |
| Always (installation) | I1, I2, I3, I4, I5 | |
Project has implementation_target: shadcn and .design/system/STACK.md exists | S1–S5 | |
| Multiple projects reference the same brand | X1 |
For each check, record the verdict (PASS/WARN/FAIL/INFO) plus a short instance-tagged message for issues.
De-duplication: if the same underlying issue is caught by multiple checks, only report it once — keep it in the most specific check.
Step 4: Calculate health score
Score per instance (100 points each):
- Each FAIL: -15 points
- Each WARN: -5 points
- Each INFO: 0 points
- Minimum: 0
Overall score: average of all instance scores.
Step 5: Display diagnostic
Render to terminal only (no file writes). One section per instance, then installation, stack, cross-instance, then issues, then summary.
Verdict glyphs: ✅ PASS · ⚠️ WARN · ❌ FAIL · ℹ️ INFO Health emoji: 90–100: 💚 · 70–89: 💛 · 50–69: 🟠 · 0–49: ❤️ Health bar: 20-char using █ and ░.
Summary lines by score band:
≥ 90— "Project is healthy. Ship it! 🚀"70–89— "Project has minor issues. Address warnings when convenient."50–69— "Project needs attention. Fix the warnings above."< 50— "Project has significant issues. Address failures first."
Important Notes
- Read-only — do NOT modify any files
- No agents — run all checks directly, this is deterministic pattern matching
- Terminal only — no file output, all results printed to terminal
- Be specific — every issue names the exact file and suggests the exact command to fix it
- Don't over-report — if the same issue is caught by multiple checks, only report it once (in the most specific check)
</process>
Security
- Security scanPassed, 100 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckinggsp/skills/gsp-doctor/SKILL.md
Popularity
- Repository stars
- 43
- Repository forks
- 4
- Repository npm downloads, 7 days
- 59
- 7-day trend score
- 32
Stars and forks are for the whole GitHub repository, not this skill alone. Downloads are for the repository's npm package. Refreshed 17 May 2026, 5 months ago.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/jubscodes/get-shit-pretty/gsp-doctor)