frontend-engineering
Skill4 starsChanged 40 days agoMITView on GitHub
Trigger: frontend engineering, frontend architecture, UI design, component design, state management, accessibility, design system, UI review, frontend docs. Guide frontend work from discovery to delivery, adapting rigor to size and risk.
13 other projects use a version of this.
Adopted by
Most recently taken up by
- juninmd/skillstheir own version, first seen Nov 8, 2026
- hamzzaqureshi/gravidextheir own version, first seen Nov 8, 2026
- rogue-dev-studio/ai-agents-roguetheir own version, first seen Nov 1, 2026
- DevelopersGlobal/ai-agent-skillstheir own version, first seen Oct 25, 2026
- PhoenixNest/AGENT-GLOBAL-BASEtheir own version, first seen Oct 20, 2026
- TylerAkins/ai-agent-skillstheir own version, first seen Oct 15, 2026
Used with
Projects with it also use observability and git-workflow.
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.
OpenAISkillobservability
Vercel Observability expert guidance — Drains (logs, traces, speed insights, web analytics), Web Analytics, Speed Insights, runtime logs, custom events, OpenTelemetry integration, and monitoring dashboards. Use when instrumenting, debugging, or optimizing application performance and user experience on Vercel.
Also in 3 projects that use this
Microsoft AzureSkillgit-workflow
Squad branching model: dev-first workflow with insiders preview channel
Also in 3 projects that use this
Other versions13 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 13
- DevelopersGlobal/ai-agent-skills★ 216same name
- magnus919/agent-skills★ 172same name
- d-padmanabhan/agent-engineering-handbook★ 17same name
- JimmyVGDY/codex-long-term-assistant-skills★ 3same name
- adtn0810/truestack★ 2same name
- juninmd/skills★ 2same name
- PhoenixNest/AGENT-GLOBAL-BASE★ 2same name
- rogue-dev-studio/ai-agents-rogue★ 2same name
- hamzzaqureshi/gravidex★ 1same name
- TylerAkins/ai-agent-skills★ 1same name
3 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 repository18 other files it gives its agents.
- api-designSkill
- changelog-maintenanceSkill
- code-mentorSkill
- create-readmeSkill
- database-designSkill
- dockerSkill
- frontend-design-anti-slopSkill
- git-commitSkill
- github-essentialsSkill
- java-junitSkill
- java-springbootSkill
- java-springboot-testingSkill
- latexSkill
- solution-designSkill
- ui-componentsSkill
- update-readmeSkill
- vault-tech-noteSkill
- web-mvcSkill
The file
We split this skill into the rules it gives an agent
9 rules. 0 are common, 9 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
9
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 28
Do not load for a tiny, isolated change: inspect the relevant context, implement it, run appropriate checks, and summarize remaining risks …
- Distinctive: line 37
Design states, not just happy paths: Consider loading, empty, error, success, disabled, permission-denied, offline, and partial-data states …
- Distinctive: line 39
Keep documentation aligned with the code: Update only affected documentation.
- Distinctive: line 40
Validate incrementally: Prefer small, testable slices of user-visible functionality over a large speculative design phase.
- Distinctive: line 41
Follow the project's conventions: Do not introduce new libraries or architectural patterns without a reason and an explicit trade-off.
- Distinctive: line 42
The client is not a trust boundary: Never put secrets or privileged authorization logic in client code.
- Distinctive: line 78
Start with the smallest useful set of documents.
- Distinctive: line 86
Read only the references relevant to the current task:
- Distinctive: line 108
Before considering a design or implementation task complete, verify the relevant items:
“Common” is measured: kept by at least 6 in 10 of the 9 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
Frontend Engineering
Help plan, design, build, and maintain frontend software with an appropriate level of rigor. Apply the skill to web frontends and, where relevant, mobile or desktop clients. Framework-agnostic: never assume a particular framework, design system, state library, or build tool; adapt examples to the actual stack.
This skill owns frontend craft. When the question is cross-cutting process (documentation level, artifact organization, requirement discovery), defer to the solution-design skill if available and keep only frontend-relevant consequences here.
Activation Contract
Load this skill when the user:
- starts a frontend project from scratch;
- plans a UI, user flow, component system, or frontend architecture;
- implements a feature spanning multiple screens or components;
- reviews frontend maintainability, UX, accessibility, performance, or testing;
- creates or updates frontend design documentation;
- aligns frontend behavior with a backend API or product requirements.
Do not load for a tiny, isolated change: inspect the relevant context, implement it, run appropriate checks, and summarize remaining risks — no full workflow.
Hard Rules
- Understand before changing code. For an existing project, inspect the repository, package manifests, routing, components, styling, tests, and existing documentation before proposing changes. Never assume a clean slate.
- Adapt to scale and risk. Choose lightweight, standard, or rigorous using complexity, user impact, accessibility needs, security/privacy, performance requirements, integrations, team size, and uncertainty — not only the number of screens.
- Avoid document inflation. Capture important decisions; create a separate file only when separation improves collaboration, review, traceability, or maintenance.
- Separate requirements from implementation. State what users need apart from the technical approach used to deliver it.
- Make assumptions visible. Distinguish confirmed requirements, assumptions, open questions, and recommendations. Never silently invent product behavior, roles, branding, or APIs.
- Design states, not just happy paths. Consider loading, empty, error, success, disabled, permission-denied, offline, and partial-data states when relevant.
- Accessibility and responsive behavior are quality requirements from the beginning, not a final polish pass.
- Keep documentation aligned with the code. Update only affected documentation; one source of truth, link instead of duplicating.
- Validate incrementally. Prefer small, testable slices of user-visible functionality over a large speculative design phase.
- Follow the project's conventions. Do not introduce new libraries or architectural patterns without a reason and an explicit trade-off.
- The client is not a trust boundary. Never put secrets or privileged authorization logic in client code; client-side validation improves usability but never replaces server-side validation.
Decision Gates
Documentation level
| Signals observed | Level | Artifacts |
| Small scope, low risk, few views/flows, one developer or rapid prototype | Lightweight | One docs/frontend/design.md (see templates/frontend-design.md) plus README if needed; combine requirements, flows, architecture, UI rules, API assumptions, accessibility, tests |
| Several features, shared components, meaningful client state, integrations, or a team | Standard | Concise design overview; add focused documents (ui-flows.md, design-system.md, frontend-architecture.md, frontend-testing.md, ADRs) only when separate ownership or frequent updates justify them |
| High-impact or regulated journeys, sensitive data, complex permissions, strict accessibility/performance targets, many teams, large application | Rigorous | Requirements traceability, design-system documentation, architecture decision records, test strategy, performance budgets, security/privacy reviews as applicable |
A project can be small but high-risk: increase rigor where risk warrants it. Reassess the level when scope changes. Reassessment triggers and consolidation criteria: references/workflow.md.
Handoff
| Question | Owns it |
| Cross-cutting process: requirements discovery, documentation level, artifact organization | solution-design skill, if available; otherwise keep the lightweight process above |
| Frontend craft: component boundaries, state, UX, accessibility, testing, performance | This skill |
| API/database contract detail | api-design / database-design skills, if available |
Execution Steps
- Establish context. Greenfield: clarify product goal, users, primary tasks, devices/browsers, brand constraints, accessibility expectations, data sensitivity, API dependencies, and constraints; ask focused questions only when they materially affect the design, otherwise record assumptions. Existing project: inspect framework versions, routes, styling strategy, component conventions, state/data fetching, API contracts, tests, and the current design system first.
- Select the documentation level from the Decision Gate above and briefly state why.
- Define user and interface requirements. User types and goals, key journeys, navigation, screen responsibilities, responsive behavior, and acceptance criteria. If access differs by role, document which actions and views are allowed, hidden, disabled, or rejected — visibility is a UX rule, not the security boundary. Wireframes/diagrams only when they reduce ambiguity; if visual design is requested, establish the output format (image, editable mockup, or implementation) before choosing.
- Design frontend architecture. Routes, component responsibilities (shared vs feature-specific), state ownership and data flow, API client behavior, forms and validation, permission-aware presentation, configuration, localization and error boundaries if appropriate. Prefer the simplest architecture that meets current requirements; no speculative frameworks, unnecessary global state, or premature abstraction. Reason with
references/frontend-architecture.md. - Define UX and visual behavior. Information hierarchy, interaction patterns, design tokens and component variants when a design system is justified, form UX, all interface states, responsive layouts, and the accessibility baseline. Extend an existing design system instead of creating a competing one; never invent brand colors or typography — place an explicit placeholder or ask. Reason with
references/ui-ux-accessibility.md. - Align API and data contracts. Confirm endpoints, methods, schemas, pagination, validation errors, and authentication expectations against the backend contract; reuse existing OpenAPI or generated clients. If the contract is unknown, label it as an assumption and never claim integration. Details:
references/frontend-architecture.md. - Plan quality from the beginning. Select checks proportionate to the project — unit/component/integration/e2e tests, accessibility (automated plus keyboard/manual), responsive/cross-browser per support targets, performance (bundle, rendering, Core Web Vitals) when relevant, visual regression only when justified. Tie acceptance criteria to observable tests or a documented manual method. Plan with
references/quality-testing.md. - Plan implementation in vertical slices. Small increments delivering a complete user-visible capability; for each, identify screens/components, API dependencies, states, acceptance criteria, and tests. Resolve high-risk unknowns early; do not require full specification of every screen before a useful first increment.
- Implement, verify, and maintain. Before editing, summarize the intended scope for significant work. Follow repository conventions, run relevant checks, and report checks not run and why. Update documentation when a material decision, interaction, contract, or architecture changes. For review: correctness, accessibility, security/privacy, responsive behavior, maintainability, performance, regression risk; distinguish verified issues from potential concerns.
Documentation Rules
- Start with the smallest useful set of documents; explain briefly why any separate artifact is needed.
- Lightweight: use
templates/frontend-design.mdas the basis for a singledocs/frontend/design.md. - Standard/rigorous: split sections only when the split improves use or maintenance. Do not duplicate complete API definitions, product requirements, or backend/database specifications — link to their canonical documents and record only frontend-relevant consequences.
- Mark sections
Not applicableonly when the omission needs explanation; otherwise omit them. - Label decisions
Confirmed,Proposed, orOpen questionwhen the status matters.
Reference Guide
Read only the references relevant to the current task:
references/workflow.md: project lifecycle, reassessment triggers, consolidation criteria.references/frontend-architecture.md: architecture, component boundaries, state, routing, API integration.references/ui-ux-accessibility.md: user flows, interface states, responsive design, accessibility, design systems.references/quality-testing.md: testing, performance, security/privacy, release checks.references/document-templates.md: suggested sections for lightweight and specialized documents.templates/frontend-design.md: concise all-in-one design document template for small projects.
Output Contract
At the end of a design or implementation task, report:
- Files created or modified (including where each documented responsibility lives if artifacts were combined).
- The documentation level chosen and why.
- The most important assumptions.
- Open questions, blockers, and pending high-impact decisions.
- Checks that were run, checks that were not run and why.
- The next increment recommended.
Completion Checklist
Before considering a design or implementation task complete, verify the relevant items:
- user goals and scope are understood;
- key flows and acceptance criteria are clear;
- component and state responsibilities are coherent;
- API dependencies and unknowns are explicit;
- loading, empty, error, success, and permission states are considered where relevant;
- responsive behavior and accessibility are addressed;
- security-sensitive behavior is not entrusted to the client alone;
- appropriate tests/checks are run, or remaining gaps are reported;
- documentation is concise, current, and not unnecessarily duplicated.
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.

