workflow-build-feature
Skill9 starsChanged 42 days agoMITView on GitHub
Build a feature end to end: spec and TDD, implement, unit tests, Playwright check, PR. Use when "build a feature" or "implement this end-to-end". Spec/TDD loop only → workflow-spec-tdd. One bug → workflow-fix-and-ship.
2 other projects use a version of this, 1 of them edited.
Adopted by
Most recently taken up by
- ProxiBlue/claude-skillstheir own version, first seen Oct 7, 2026
- kensaurus/skillsedited, first seen Oct 1, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 1
- kensaurus/skills★ 999% same
Wrote their own version · 1
- ProxiBlue/claude-skills★ 18same name
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 repository165 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/approved-plan-execution.mdcCursor rule
- .cursor/rules/skill-workflows.mdcCursor rule
- .cursor/rules/verification-before-completion.mdcCursor rule
- llms.txtllms.txt
- audit-accessibilitySkill
- audit-agent-speedSkill
- audit-analyticsSkill
- audit-auth-flowsSkill
- audit-backend-architectureSkill
- audit-bundle-sizeSkill
- audit-cicdSkill
- audit-codemod-safetySkill
- audit-code-qualitySkill
- audit-code-reviewSkill
- audit-db-schemaSkill
- audit-doctrineSkill
- audit-env-paritySkill
- audit-fe-apiSkill
- audit-gate-logicSkill
- audit-i18nSkill
- audit-infra-costSkill
- audit-langfuse-llmSkill
- audit-llm-securitySkill
- audit-monetization-iapSkill
- audit-payment-systemSkill
- audit-performanceSkill
- audit-realworldSkill
- audit-registry-listingSkill
- audit-resilienceSkill
- audit-responsiveSkill
- audit-securitySkill
- audit-skill-conflictsSkill
- audit-ui-statesSkill
- audit-uiux-design-systemSkill
- audit-ux-journeysSkill
- audit-uxSkill
- backend-db-performanceSkill
- backend-error-handlingSkill
- backend-observabilitySkill
- backend-patternsSkill
- backend-realtimeSkill
- burndown-fullSkill
- complete-everythingSkill
- data-pipelineSkill
- data-visualizationSkill
- debug-errorSkill
- debug-fe-be-integrationSkill
- debug-sentry-monitorSkill
- deploy-npmSkill
- deploy-verifySkill
- design-apiSkill
- design-canvasSkill
- design-emailSkill
- design-frontendSkill
- design-generative-artSkill
- design-mobile-firstSkill
- design-motionSkill
- design-prdSkill
- design-systemSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 65
When the ask is loosely specified, read the relevant routes, schema, tests, and sibling features broadly before writing the spec, so it is …
- Very specific: line 97
Before any browser action, apply protocol-browser-anti-stall rules.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
workflow-build-feature — End-to-End Feature Build
Degree of freedom: MIXED. How to implement Phase 2 [HIGH freedom];
phase order and "no implementation before RED" [LOW freedom — run exactly].
One command that runs the full feature loop. Don't skip phases — each gate catches a class of defects the next phase can't.
How to reason
- Contract — what observable behavior is in / out of scope?
- Sequence — which phase is next, and what would skipping it hide?
- Prove — RED was real; GREEN and smoke cover the contract
- Handoff — PR evidence is complete; no silent "looks done"
Worked example
Contract: "add a forgot-password link" — email reset, not a magic-link login rewrite. Sequence: spec + RED on the request-reset handler before any UI. Prove: RED failed on missing route; GREEN + Playwright submits the form and sees the check-email state. Handoff: PR links the spec, the RED/GREEN commands, and the smoke screenshot.
Self-critique before reporting
- Phases intact — none of spec / RED / unit / smoke / PR were skipped
- RED was real — the failing test existed before implementation
- Full-stack — if data was touched, the migration was applied and verified
- Right owner — one named bug →
workflow-fix-and-ship; spec loop only →workflow-spec-tdd
Phase sequence [LOW freedom — run exactly]
1. SPEC → workflow-spec-tdd (contract, plan, RED test)
2. BUILD → implement against the spec (GREEN)
3. UNIT TEST → test-unit (edge cases, mocks, coverage)
4. SMOKE → test-playwright (drive the live app as a real user, fix pain points)
5. PR → workflow-pr (create PR with test evidence)The turn ends at the Done criteria or at a real gate — a decision only the user can make, a missing credential, a failing service. A phase summary is a progress note, not a stopping point.
Phase 1: Spec (read workflow-spec-tdd) [LOW freedom — run exactly]
Read the
workflow-spec-tddskill and follow it.
Key outputs:
- Written spec (what it does, what it doesn't do, acceptance criteria)
- File-mapped implementation plan (which files to create/edit)
- A RED failing test that encodes the acceptance criteria
When the ask is loosely specified, read the relevant routes, schema, tests, and sibling features broadly before writing the spec, so it is grounded in what the repo does rather than the first plausible reading.
Do not write implementation code until the spec and RED test exist.
Phase 2: Build [HIGH freedom]
Implement against the spec. Rules:
- One logical change per commit
- Match the file patterns already in the repo — no new abstractions unless the spec calls for them
- Full-stack: if the feature reads/writes data, verify the DB migration is applied before the frontend ships (see full-stack-ship-discipline rule)
- Scope stays where the spec put it. A pre-existing bug you find on the way is a follow-up in the PR description, not part of this change. Scratch scripts and one-off checks stay out of the repo. Edit files surgically — change the lines the feature needs, do not rewrite the file around them.
Phase 3: Unit tests (read test-unit) [LOW freedom — hand off]
Read the
test-unitskill and follow it.
Focus on:
- The new code's happy path, error paths, and edge cases
- Any shared utility functions introduced
Phase 4: Smoke test (read test-playwright) [LOW freedom — hand off]
Read the
test-playwrightskill and follow it.
Drive the live app through the new feature as a real user. Fix pain points inline (full-stack: UI + API + DB). Capture screenshots as evidence.
Before any browser action, apply protocol-browser-anti-stall rules.
Phase 5: PR (read workflow-pr) [LOW freedom — hand off]
Read the
workflow-prskill and follow it.
Include in the PR description:
- Link to the spec / acceptance criteria
- Screenshots or screen recording from Phase 4
- Migration notes if schema changed
Done criteria [LOW freedom — do not skip]
- Spec written and agreed
- RED test exists and was failing before implementation
- Implementation makes test GREEN
- Unit tests cover happy path + at least one error path
- Smoke test drove the live app and found no blockers
- DB migration applied and verified (if applicable)
- PR open with description and evidence
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.

