frontend-ui-verification
Use when a Clipboard frontend task involves a new UI idea, Figma/design URL, screenshot, redesign UI, design-to-code implementation, Storybook visual validation, Playwright/browser screenshot, animation/motion visual proof, or preventing design-system and visual-fidelity drift in admin or mobile UI work.
What this skill does
# Frontend UI Verification
Use this skill for UI work where visual fidelity, design-system reuse, responsive behavior, or human review matters. The goal is to make frontend visual changes cheap to inspect before real integration makes feedback slow and expensive.
## Core Rule
Code is the source of truth for available components, story patterns, and design-system usage. Design references are the source of truth for intended appearance and behavior. Do not invent components, wrappers, or design tokens until you have searched the repo and checked the relevant design/reference source.
Keep this as a cheap visual layer. Use repo rules and agent judgment for routine React implementation details. Be prescriptive only where agents commonly drift: choosing the reference source, finding existing components, creating an isolated Storybook surface, managing the Storybook server lifecycle, opening the exact canvas URL, capturing browser evidence, classifying console findings, and deleting temporary checkpoint stories.
## Required Workflow
1. **Identify the surface**
- Determine repo and surface: admin web, mobile web/app, shared package, or another frontend.
- Identify the strongest available reference source:
- Figma/design URL: fetch screenshot, design context, variables/tokens, metadata, and design-system/Code Connect mappings when available. If the URL is inaccessible, falls behind login, or returns an error, ask for an export/screenshot or treat the work as screenshot/written-idea based; do not claim exact Figma fidelity without design evidence.
- Screenshot/image: use it as the visual reference, and call out missing states, viewport, or interaction details.
- Written idea only: do not block. Search existing components and stories, choose the closest Clipboard pattern, build a Storybook checkpoint, and ask for confirmation before deeper integration if the visual interpretation is subjective.
- No visual reference: derive from nearby product patterns and design-system components, then verify the proposed interpretation in Storybook before product integration.
- Classify visible UI work as design-first: implement the smallest inspectable surface, verify it visually, then move to product integration.
2. **Find the existing pattern first**
- Search for existing components, stories, hooks, fixtures, and nearby feature folders before creating anything.
- In `cbh-admin-frontend`, prefer `src/appV2/redesign/**` and Berlin components from `src/appV2/redesign/components/**`.
- Reuse existing Clipboard components, tokens, layouts, decorators, and story patterns whenever they satisfy the job.
- Add a new component only when existing components cannot express the needed UI without awkward composition.
- Create a generic shared component only when it is reusable across multiple surfaces and matches the established component API style. Otherwise keep it feature-local.
- Keep reusable components loosely coupled: pass primitive/domain props, not raw API response objects.
3. **Build an isolated verification surface**
- For meaningful visual work, build or update the visual slice in Storybook before product integration.
- Use a temporary story when the goal is to validate a design slice before wiring backend/auth/routing. It is acceptable for this story to contain the full proposed UI composition with fixtures while the real product code is still being shaped.
- Iterate inside Storybook: apply visual feedback, retake screenshots, and keep refining until the Storybook surface matches the reference or the human confirms the interpretation.
- After the UI is integrated, delete temporary checkpoint stories. Keep only stories that are useful permanent documentation: generic components, reusable compositions, or important product states.
- Cover the states the user would naturally inspect: default, loading, empty, error, disabled, long text, dense data, permission/flag variants, and desktop/mobile variants when the design differs.
- For drawers, menus, popovers, steppers, tabs, forms, and animated UI, make the story render deterministic pre-action states such as closed, opened, selected, invalid, in-flight, and settled.
- Mock data at the story boundary. Do not force live API, auth, or route setup just to inspect visual composition.
4. **Verify with browser evidence**
- Run the repo Storybook server as the primary visual verification surface.
- Start long-running Storybook servers in a process you can stop. In CLI or prompt-command validation, redirect logs to `/tmp`, record the server PID, and stop it before the final report. Do not leave a foreground dev server blocking screenshot capture, cleanup, or the final response.
- Open the exact story in a browser-controlled surface. Prefer the active Browser/Chrome tool when available; otherwise use Playwright against the Storybook URL.
- If the Storybook UI shell is noisy, open the canvas URL directly with `iframe.html?viewMode=story&id=<story-id>`.
- Use native Storybook test/a11y or component-level Playwright scripts only when the repo already exposes them. Do not invent a new harness for this layer.
- Check console errors, layout overflow, text clipping, loading/empty states, keyboard/interaction cues, and desktop/mobile viewport sizes.
- Classify console findings as story-introduced, provider/decorator gaps, or pre-existing/global Storybook warnings. Fix story-introduced errors before proceeding; report global warnings separately instead of hiding them.
- If motion or animation is part of the change, verify the motion in Storybook with a repeatable or controllable story, then capture evidence after the animation starts and after it settles.
- Capture screenshots when visual fidelity is part of the claim. For Figma work, compare the Storybook screenshot against the Figma screenshot element by element.
- If the design is non-trivial or the user needs sign-off, share the Storybook URL and screenshot summary, then wait for confirmation before deeper integration.
5. **Do not give up on visual access early**
- Run the helper's `inspect` command first when the repo command, build command, or story URL shape is unclear.
- If the Storybook dev server fails, try the repo's Storybook build command and inspect the built Storybook when practical.
- If port `6006` is busy, use the alternate port Storybook prints and report that exact URL.
- If the story fails because of providers, routing, API data, or feature context, fix the story with decorators or fixture props instead of switching to product integration.
- If one browser tool fails, try another available browser path: Browser plugin, Chrome, Playwright CLI/script, or the agent's equivalent.
- Stop only after at least two reasonable Storybook access paths fail, and report the exact command, URL, error, and what evidence is missing.
6. **Ask for human confirmation at the right time**
- Ask before choosing between multiple plausible visual interpretations.
- Ask after a temporary Storybook checkpoint when the user requested exact visual match, when the reference is subjective, or when the design changes layout density/hierarchy.
- Ask before adding a new generic component unless there are at least two concrete reuse sites or an existing design-system gap.
- Do not ask for routine wiring details once the Storybook surface clearly matches existing patterns.
7. **Hand off after the visual surface is sound**
- If the UI is simple and maps directly to existing components, proceed to integration after visual inspection.
- Treat full product workflow checks as a separate post-verification layer.
- Do not expand this skill into full-flow Playwright or product-environment work.
8. **Final verification report**
- List the reference source used: Figma/Notion/screenshot/written idea/story/code path.
- List the Storybook URL checked.
- List viewporRelated in Design
contribute
IncludedLocal-only OSS contribution command center. Auto-refreshes the user's in-flight PR and issue state on invoke so conversations start with full context — no need to brief Claude on what's in flight. Helps the user find issues to contribute to on GitHub, builds per-repo dossiers of what each upstream expects (CLA, DCO, branch convention, AI policy, draft-first, review bots, issue templates), runs deterministic gates before any external action so AI-assisted contributions don't reach maintainers as slop. State is markdown-only: candidate files at ~/.contribute-system/candidates/, repo dossiers at ~/.contribute-system/research/, append-only event log at ~/.contribute-system/log.jsonl. No database, no cloud calls. Use when the user asks about their PRs / issues / contributions, wants to find new work to take on, claim an issue, build/refresh a repo's dossier, or draft a Design Issue or PR. Trigger with "/contribute", "what's my PR status", "find a contribution", "claim issue X", "draft a Design Issue for Y", "refresh dossier for Z".
architectural-analysis
IncludedUser-triggered deep architectural analysis of a codebase or scoped subtree across eight modes — information architecture, data flow, integration points, UI surfaces, interaction patterns, data model, control flow, and failure modes. This skill should be used when the user asks to "diagram this codebase," "map the architecture," "show the data flow," "give me an ERD," "trace control flow," "find the integration points," "verify the layout pattern," "audit the UX architecture," or any similar request whose primary deliverable is mermaid diagrams plus cited reports under docs/architecture/. Dispatches haiku/sonnet sub-agents in parallel for per-mode exploration, then verifies every citation mechanically before any node lands in a diagram. Not for one-off prose explanations of code (use code-explanation) or for high-level system design from scratch (use system-design).
mcp
IncludedModel Context Protocol (MCP) server development and tool management. Languages: Python, TypeScript. Capabilities: build MCP servers, integrate external APIs, discover/execute MCP tools, manage multi-server configs, design agent-centric tools. Actions: create, build, integrate, discover, execute, configure MCP servers/tools. Keywords: MCP, Model Context Protocol, MCP server, MCP tool, stdio transport, SSE transport, tool discovery, resource provider, prompt template, external API integration, Gemini CLI MCP, Claude MCP, agent tools, tool execution, server config. Use when: building MCP servers, integrating external APIs as MCP tools, discovering available MCP tools, executing MCP capabilities, configuring multi-server setups, designing tools for AI agents.
react-native-skia
IncludedDesign, build, debug, and optimise high-polish animated graphics in React Native or Expo using @shopify/react-native-skia, Reanimated, and Gesture Handler. Use when the user wants canvas-driven UI, shaders, paths, rich text, image filters, sprite fields, Skottie, video frames, snapshots, web CanvasKit setup, or performance tuning for custom motion-heavy elements such as loaders, hero art, cards, charts, progress indicators, particle systems, or gesture-driven surfaces. Also use when the user asks for fluid, glow, glass, blob, parallax, 60fps/120fps, or GPU-friendly animated effects in React Native, even if they do not explicitly say "Skia". Do not use for ordinary form/layout work with standard views.
plaid
IncludedProduct Led AI Development — guides founders from idea to launched product. Six capabilities: Idea (discover a product idea), Validate (pressure-test the idea against fatal flaws, problem reality, competition, and 2-week MVP feasibility), Plan (vision intake + document generation), Design (translate image references into a design.md spec), Launch (go-to-market strategy), and Build (roadmap execution). Use when someone says "PLAID", "plaid idea", "help me find an idea", "product idea", "idea from my business", "idea from my expertise", "plaid validate", "validate my idea", "pressure-test", "is this idea good", "find fatal flaws", "validate the problem", "plan a product", "define my vision", "generate a PRD", "product strategy", "plaid design", "design from image", "translate image to design", "create design.md", "extract design tokens", "plaid launch", "go-to-market", "launch plan", "GTM strategy", "launch playbook", "plaid build", "build the app", "start building", or "execute the roadmap".
nextjs-framer-motion-animations
IncludedAdds production-safe Motion for React or Framer Motion animations to Next.js apps, including reveal, hover and tap micro-interactions, whileInView, stagger, AnimatePresence, layout and layoutId transitions, reorder, scroll-linked UI, and lightweight route-content transitions. Use when the user asks to add, refactor, or debug Motion or Framer Motion in App Router or Pages Router codebases, especially around server/client boundaries, reduced motion, LazyMotion, bundle size, hydration, or route transitions. Avoid for GSAP-style timelines, WebGL or 3D scenes, heavy scroll storytelling, or CSS-only effects unless Motion is explicitly requested.