Mobbin Claude Rodeo — AI skill for Claude Code
Claude Code skill: find real-app UI references on Mobbin and build to them faithfully — no LLM drift, no stale specs.
How to install Mobbin Claude Rodeo
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open b1rdmania/mobbin-claude-rodeo and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Mobbin Claude Rodeo does
Claude Code skill: find real-app UI references on Mobbin and build to them faithfully — no LLM drift, no stale specs.
Alternatives in AI
- Opencodex — Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama… 12.3k ★
- Manifest — Real-time cost observability for OpenClaw agents — track tokens, costs, messages, and model usage 4.1k ★
- Claude Usage Tracker — Native macOS menu bar app for tracking Claude AI usage limits in real-time 2.2k ★
README
Mobbin Claude Rodeo 🤠🎯
**Reference guardrails for Claude — find a real UI on [Mobbin](https://mobbin.com), build it *exactly*, and never let the spec go stale.**
What This Does
Claude is great at *finding* strong UI references and bad at *honouring* them. Two failures recur:
- Approximation. You find a sharp sidebar — "compact left rail, 220px, 5 items." Claude builds a Claude-flavoured impression of it instead: generic spacing, reordered items, an invented active-state, "helpful" icons. The reference is gone.
- Stale specs. You capture it well, build it faithfully — then the IA changes. The code moves on, the spec doesn't, and now nobody can tell deliberate change from drift. Every future "does this match the design?" is unanswerable.
**This skill** turns a Mobbin reference into something *load-bearing*: extracted as exact values, built without drift, and kept honest as the product evolves.
How It Works
flowchart LR
A["1 · SEARCH
mine Mobbin
by job + archetype"] --> B["2 · EXTRACT
screen → P## spec
of exact values"]
B --> C["3 · BUILD
strict preservation
copy, don't improve"]
C --> D["4 · AUDIT
code vs spec
drift list"]
D -->|"drift → fix the code"| C
D -->|"deliberate change → update the spec"| B
The loop is the point. A drift sends you back to the **code**; a deliberate product change sends you back to the **spec** — so the reference never silently rots.
| Phase | What happens |
|---|---|
| 1 · Search | MCP-first (mcp__mobbin__search_screens / search_flows), manual-screenshot fallback. Query the job + archetype, not the aesthetic. Stop at 2–3 convergent precedents. |
| 2 · Extract | Turn the chosen screen into a numbered P## spec block of exact values (px, token, item order, presence). Write it into your DESIGN.md. Adjectives aren't extraction. |
| 3 · Build | Copy spec values exactly. No "improvements". Build th |
Related Skills
Grep Jsonl
Inspect Mori's ~/.mori/logs/mori-YYYY-MM-DD.jsonl event log to diagnose user-reported issues — find errors, tr
Claude Skill Web Clone
Methodology for faithfully cloning any website (static / React / WebGL) — without copying AI-hallucinated code
Fallow Skills
Agent skills for fallow, codebase intelligence for TypeScript and JavaScript. Teaches AI agents how to find un
UX Lint
Deterministic regex-based linter for AI fingerprints. No LLM, no API, no network. CI-friendly — exits non-zero
Bead Sweep
Scan open beads for stale trackers already implemented — deterministic checks then LLM verification.
AI Keyword Research Agent
AI Keyword Research Automation — give your AI coding assistant one job: find N real keyword opportunities for
Related Agents
Review Workflow Consistency
Reviews .claude/ workflow machinery for cross-file consistency: phantom references, mismatched threshold table
Visual QA
Read-heavy visual QA role for real game/app states. Captures or inspects mobile/desktop evidence, flags clippi
Docs Drift Review
Given a session's commits or diff, finds the docs that the change made wrong — stale paths and symbols, checkl