frontend
Skill1 starsChanged 6 days agoMITView on GitHub
Frontend engineering guidance - component architecture, state management, CSS/styling architecture, forms and validation, routing, responsive/mobile design, frontend performance, SEO/AI-visibility, and accessibility. Use for any UI/component/frontend-structure question, including what to test on the frontend (see codexpiator:testing-qa for general testing strategy/TDD discipline instead).
54 other projects use a version of this.
Adopted by
Most recently taken up by
- nomorehappypath/nomorehappypaththeir own version, first seen Nov 15, 2026
- haxsd/haxsd-byoktheir own version, first seen Nov 15, 2026
- theopenco/llmgatewaytheir own version, first seen Nov 15, 2026
- speakeasy-api/gramtheir own version, first seen Nov 13, 2026
- Mazalucas/El-DT-Lightweight-Armytheir own version, first seen Nov 13, 2026
- notque/vexjoy-agenttheir own version, first seen Nov 10, 2026
Used with
Projects with it also use programming, data and 6 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.
DevNewbie1826Skillprogramming
MUST USE for work on Python, Rust, TypeScript, JavaScript, or Go source files and project manifests, and whenever a test or spec file is written, reshaped, reviewed, or deleted. Triggers: write or edit code; add, change, or delete a test; test authoring gate; junk tests; new project; Go HTTP service (gin lane); Bubble Tea TUI; CJK IME; Connect-Go RPC; sqlc or pgx; branded IDs; exhaustive matching; unsafe Rust; Miri; oversized file; refactoring; logging, log levels, structured logging, or observability; arena, allocator, bumpalo, const fn, const generics, compile-time programming, zero-allocation APIs, bitfields, repr, scope guards, errdefer-style cleanup, Zig-like patterns, zero-copy parsing, packed data, or FFI; Rust public API design or macros.
Also in 3 projects that use this
launch52-aiSkilldata
Generate data layer code (Freezed DTOs, repository implementations, data sources) from feature specifications or API references. Creates the bridge between domain entities and external data sources. Use after /domain to implement the data layer.
Also in 4 projects that use this
launch52-aiSkilldomain
Generate domain layer code (entities, enums, repository interfaces) from feature specifications. Creates pure Dart business logic with no external dependencies. Use after /feature-init to fill in the domain layer.
Also in 3 projects that use this
mijunriSkilldatabase
数据库脚本与操作规范。脚本放在 db_scripts 目录,遵循查询优化、事务、安全等实践。
Also in 3 projects that use this
BlockSkilldebugging
Debugging guidelines and best practices
Also in 6 projects that use this
MicrosoftSkillhelp
Entry point for the MedImageInsight plugin, covering the getting-started walkthrough, how the skills work, what MI2 is, and troubleshooting.
Also in 4 projects that use this
Other versions54 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 54
- langwatch/langwatch★ 5ksame name
- atopile/atopile★ 4ksame name
- leookun/cursor-byok★ 3.2ksame name
- theopenco/llmgateway★ 1.7ksame name
- notque/vexjoy-agent★ 450same name
- notque/claude-code-toolkit★ 448same name
- AlexPEClub/ai-coding-starter-kit★ 382same name
- OyadotAI/oya-browser★ 349same name
- speakeasy-api/gram★ 278same name
- ThePrimeagen/skills★ 234same name
44 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 repository9 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- ai-integrationSkill
- architectureSkill
- backendSkill
- devopsSkill
- dx-gitSkill
- securitySkill
- routerSkill
- testing-qaSkill
The file
We split this skill into the rules it gives an agent
7 rules. 0 are common, 3 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 48 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 9
Read the specific file below for the concern at hand rather than trying to hold all of it in one place.
- Very specific: line 29
If the task is primarily about visual design quality, motion, a broad polish pass, or auditing an existing UI, check …
- Very specific: line 79
Before dispatching the selection subagent, call AskUserQuestion to cover (skip only what's already been stated explicitly or is genuinely …
- Very specific: line 107
Don't bulk-download the whole collection.
- Distinctive: line 121
Before considering frontend work done:
- Distinctive: line 128
DEFAULT patterns that fit, proposed the PROPOSE ones that fit.
- Very specific: line 145
If the project is still using a placeholder/default favicon or brand images (per seo-and-launch-checklist.md), remind the user at the end …
“Common” is measured: kept by at least 6 in 10 of the 48 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
Codexpiator Frontend
Single-purpose reference files for frontend engineering decisions. Read the specific file below for the concern at hand rather than trying to hold all of it in one place.
| File | Read this when... |
component-architecture.md | Structuring components, deciding what to extract, naming, folder layout |
state-management.md | Choosing where state lives, local vs global, server vs client state |
styling-and-css.md | Choosing a styling approach, design tokens, theming, responsive strategy |
forms-and-validation.md | Building forms, validation timing, accessible error handling |
routing-and-navigation.md | Route structure, code-splitting by route, auth-gated routes |
responsive-and-mobile.md | Breakpoints, touch targets, mobile web specifics |
frontend-performance.md | Bundle size, lazy loading, images, layout shift |
frontend-testing.md | What to unit/integration/E2E test on the frontend |
accessibility.md | Semantic HTML, ARIA, keyboard nav, contrast, focus management |
ux-polish-patterns.md | Building or reviewing any UI: skeletons, optimistic updates, undo toasts, command menu, shortcut hints, SWR, pre-rendering, blurhash, view transitions, swipe sheets, interruptible animations, predictive back, hit stop, safe areas, OTP autofill, tabular numbers, balanced headlines, continuous corners, safe triangles |
seo-and-launch-checklist.md | Pre-launch pass: metadata, SEO tags, legal pages, conversion essentials |
seo-strategy-and-ai-visibility.md | Deep SEO strategy + GEO: ranking in Google/Bing/DuckDuckGo and being cited by AI answer engines |
Before any visually-led answer
If the task is primarily about visual design quality, motion, a broad
polish pass, or auditing an existing UI, check ${CLAUDE_PLUGIN_ROOT}/shared/external-skills-registry.md
first — a specialized external skill may be a better fit than this
skill's own guidance, and that file has the procedure for checking.
This is not optional for design-heavy work: attempt to consult the
relevant external design skill(s) before finalizing an answer, not
only as a fallback when you happen to remember to check.
Apply UX polish patterns by default
For any UI work, read ux-polish-patterns.md and apply every
DEFAULT pattern that fits the screen without being asked (skeleton
loaders, tabular numbers, balanced headlines, safe-area insets,
one-time-code autofill, interruptible animations, shortcut hints, undo
toasts, stale-while-revalidate, safe triangles on hover flyouts,
predictive-back-friendly navigation). PROPOSE the others in one line
when the situation fits (optimistic updates, command menu, view
transitions, swipe-to-dismiss sheets, pre-rendering, blurhash, continuous
corners), and use AskUserQuestion when several apply. Hit stop only for
game-like UI. When the user names one of these, implement it. In the
final message, say which you applied and which you skipped and why.
Verify visually with Playwright
For any non-trivial frontend change, actually look at the rendered result before calling it done — use Playwright (via its MCP tools when available) to navigate to the page, interact with it, and take a screenshot/snapshot rather than reasoning about the DOM only from source code. If Playwright isn't available in the environment, strongly recommend the user install it rather than skipping visual verification silently.
Never commit anything Playwright creates — screenshots, traces,
test-results/, playwright-report/, blob-report/, or any ad-hoc
screenshot file taken during manual verification. These are
throwaway verification artifacts, not project deliverables; confirm
they're gitignored (see
codexpiator-dx-git/repo-hygiene-and-onboarding.md) and never git
add one even if it isn't.
Reference material: awesome-design-md
For design-heavy frontend work, pull relevant reference docs from https://github.com/voltagent/awesome-design-md — a curated collection of design-principle documents — rather than guessing from memory when an authoritative written reference is one fetch away.
Gather the user's actual preferences before selecting anything.
Which docs are worth fetching depends entirely on taste and context
that isn't in the repo or a one-line prompt — don't infer it silently.
Before dispatching the selection subagent, call AskUserQuestion to
cover (skip only what's already been stated explicitly or is genuinely
fixed by an existing design system in the repo):
- Visual style/aesthetic — e.g. minimalist, editorial, brutalist, glassmorphism, playful, corporate/professional, dark-tech — offer a real spread of distinct directions, not just light/dark.
- Language/locale — the UI's primary language(s), and whether i18n/RTL support is needed.
- Formatting & typography conventions — type scale approach, heading hierarchy style, content density (spacious vs. compact).
- Animations & motion — none/minimal, subtle micro-interactions,
or expressive/bold motion (bridges to
${CLAUDE_PLUGIN_ROOT}/shared/external-skills-registry.md'sdesign-motion-principlesentry for the deeper implementation). - Colors — a specific palette/brand colors if they have one, or a direction (warm/cool/neutral, monochrome vs. saturated) if not.
- Layout patterns — e.g. classic marketing sections, dashboard/ data-dense, asymmetric/editorial grid, card-heavy vs. content-first.
- Fonts — a specific typeface if they have one, or a style direction (geometric sans, humanist sans, serif-forward) if not.
For each, offer a genuinely wide range of concrete options rather than two extremes, and mark one or two as recommended defaults so the user can pick quickly without having to have a fully-formed opinion on every axis. Only after this is settled should the docs to fetch be decided — the preferences gathered here are what the selection subagent should match against.
Don't bulk-download the whole collection. Use the Task/Agent tool to
launch the codexpiator-design-ref-picker agent (model per
${CLAUDE_PLUGIN_ROOT}/shared/model-routing.md), which browses the repo's index and picks the docs
that actually match the current project's scope and the preferences
just gathered (stack, kind of UI, and the stated style/motion/color/
layout/font direction); save only those selected docs to
.codexpiator/design-refs/ (see
${CLAUDE_PLUGIN_ROOT}/shared/long-task-memory-and-superpowers.md for the shared
.codexpiator/ directory convention — gitignore it immediately if
it's the first thing written there). Read from that local cache on
later reference within the same project instead of re-fetching.
Mandatory pre-completion gate for any frontend change
Before considering frontend work done:
- Consult every relevant file in this skill (
component- architecture.md,state-management.md,styling-and-css.md, etc. — whichever apply to the change) plus any applicable external design skill from${CLAUDE_PLUGIN_ROOT}/shared/external-skills-registry.md, not just the first one that seems relevant. 1b. Checkux-polish-patterns.mdagainst the change: applied the DEFAULT patterns that fit, proposed the PROPOSE ones that fit. - Run Prettier (or the project's configured formatter) and a type check and resolve everything they flag. Never report frontend work complete with unformatted code or type errors outstanding.
- Confirm new components/pages follow the folder-per-component
colocation convention in
component-architecture.md. - Visually verify with Playwright per the section above when the change is non-trivial — don't rely on reading code alone for something meant to be seen and used.
- Write and run real tests for whatever was created or changed —
not just the happy path, cover edge cases too (empty/loading/error
states, boundary inputs, keyboard-only interaction where relevant)
— and run the full test suite before calling the work done, not
just the new test file in isolation. See
frontend-testing.mdandcodexpiator-testing-qa/tdd-workflow.md. - Suggest running
/codexpiator:auditif the change was significant — see${CLAUDE_PLUGIN_ROOT}/shared/collaboration-and-audit-practice.md. - If the project is still using a placeholder/default favicon or
brand images (per
seo-and-launch-checklist.md), remind the user at the end of the work to provide their own custom favicon/images — don't ship or silently leave a generic default in place without flagging it.
Also see that same shared file for when to ask a clarifying question
(AskUserQuestion when available) instead of guessing.
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.

