ship— agent skill

You are in FULLY AUTONOMOUS MODE. Zero questions. Just build.

TASK: $ARGUMENTS

RULES:

• Do NOT ask the user anything. Decide and move.
• If you're unsure between two approaches, pick the simpler one.
• If a dependency is missing, install it.
• If tests don't exist, write them.
• If something breaks, fix it — don't report it, fix it.

=== PRE-BUILD: VALIDATION GATE ===

Before writing any feature code, validate the project foundation. This prevents wasting iterations on lint, platform, and config issues.

1. STATIC ANALYSIS:

• Flutter: Run flutter analyze and dart fix --apply. Fix all errors/warnings.
• Node.js: Run tsc --noEmit or the project's type-check command.
• Run the project's linter if configured.

1. PLATFORM CHECKS (Flutter):

• Scan for unguarded dart:io imports in web-reachable code.
• Verify Firebase init handles web vs native.
• Verify push notification setup is platform-guarded.

1. DEPENDENCY CHECK:

• Run flutter pub get or npm install.
• Fix version conflicts or missing packages.

1. RULES/CONFIG CHECK (if Firebase):

• Cross-check firestore.rules paths against collections used in code.
• Cross-check storage.rules against upload paths.
• Verify indexes exist for compound queries.

1. DOCKER/INFRASTRUCTURE CHECK (if docker-compose.yml or Dockerfile exists):

• Verify image references use full registry paths (e.g., ghcr.io/org/image, not bare org/image)
• Verify volume mounts: target paths exist in container, writable dirs have correct permissions
• Verify config files are mounted where the application actually reads them (check default paths, not just env vars)
• Run bash -n on all .sh scripts to catch syntax errors
• Check for macOS vs Linux portability issues: sed -i (needs '' on macOS), readlink -f (not on macOS), date flags
• Verify $SCRIPT_DIR / path references still resolve correctly after any directory reorganization

Fix everything found. Commit: "chore: pre-build validation fixes" If clean, skip the commit and proceed.

=== PER-SCREEN QUALITY CHECKLIST (CRITICAL — learned from metrics analysis) ===

Every screen you create or modify MUST satisfy these before committing. Applying these at creation time prevents 46+ retrofit commits later:

a) A11Y: Every Icon/Image has semanticLabel. Every interactive element has tooltip or Semantics label. All tap targets >= 48x48dp. Text uses theme styles. b) DESIGN TOKENS: Colors from ColorScheme only (zero hardcoded Color(0xFF...)). TextStyles from TextTheme only. Border radii from constants. Spacing on grid. Fees/rates/limits from admin config, not hardcoded. c) ASYNC SAFETY: Every async operation in StatefulWidget checks if (!mounted) return; before calling setState. d) SCALABILITY: Every Firestore query has .limit(). Batch writes for multi-doc operations. Idempotent Cloud Function triggers. e) STRUCTURAL HEALTH: If any file exceeds 500 lines, decompose it into domain-specific modules. Do not let monolithic services grow across iterations.

MONOLITH DECOMPOSITION GATE (learned from Recipe AI recall — 69 modifications to 3 monolithic files):

Before adding ANY feature code to a file that exceeds 500 lines:

1. STOP. Do not add the feature to the monolithic file.
2. Extract the relevant section into its own file first (widget, service, mixin, module).
3. Verify the extraction works (tests pass, build succeeds).
4. Commit the extraction: "refactor: extract [component] from [monolith]"
5. THEN implement the feature in the newly extracted file.

This is NOT optional. "Flag and plan" does not work — Recipe AI flagged analyze_screen.dart (6,732 lines) but never decomposed it, resulting in 21 modifications and a 62% fix-commit rate. Decompose BEFORE building, not after.

MINIMUM TEST REQUIREMENT (learned from Recipe AI — M8: Test Coverage = 0.00):

Every iteration that adds new functionality must include:

• Backend: At least 2 tests per new API endpoint (happy path + error case).
• Frontend: At least 1 widget/component test per new screen.
• If ZERO tests exist: set up test framework + 3-5 smoke tests first.

A feature is not complete until its tests exist and pass.

=== CO-COMMIT RULES (CRITICAL — learned from recall analysis) ===

These rules apply to EVERY iteration. Violating them is the #1 source of rework:

a) FIRESTORE RULES: When adding or modifying a Firestore collection, update firestore.rules in the SAME commit. Never commit feature code without rules. b) SERVER-SIDE VALIDATION: When adding client-side business logic (credit checks, eligibility, permissions), wire up server-side enforcement in the SAME iteration. c) MODEL SERIALIZATION: When a Cloud Function writes new fields, update the client model (fields, fromMap/toMap, copyWith) in the SAME commit. d) CLOUD FUNCTION TRIGGERS: When changing document structure, verify triggers still match in the same commit. e) SHARED CONFIGURATION (learned from ollama-server recall — 3 rework commits from duplicated defaults): When 2+ files reference the same configurable value (model name, base URL, API key, timeout, port), extract to a shared config module. Never hardcode the same default in multiple files. Create src/config.ts (or equivalent) with all shared defaults and env variable overrides centralized. Duplicated config is the #1 source of co-change rework in backend projects.

=== ITERATION 1: MAKE IT EXIST ===

• Build the simplest version that works.
• Co-commit Firestore rules, server validation, and model fields with features.
• Run tests/build to verify.
• Fix anything broken.
• Keep commits incremental — if touching 15+ files, split into logical commits.
• Commit: "feat: initial implementation"

=== ITERATION 2: MAKE IT SOLID ===

• Add error handling for real failure modes (not hypotheticals).
• Add or fix tests for core behavior.
• Verify all server-side validation is wired (not just client-side checks).
• Run full validation — fix until green.
• Commit: "fix: harden implementation"

=== ITERATION 3: DOMAIN ANALYSIS GATE ===

Run the /analyze skill scoped to everything you built or changed. Include all analysis phases: consistency audit, server-side validation wiring, model-to-Cloud-Function field completeness, Firebase rules, and platform compatibility.

Scope: All features/files touched across iterations 1-2. Depth: Full analysis (all phases of /analyze). Action: FIX everything rated Critical or Warning. Re-run affected checks to confirm. Commit: "fix: resolve domain analysis issues"

=== ITERATION 4: FINAL PASS (only if needed) ===

• Only run if the analysis gate found issues that required fixes.
• Re-validate everything — tests, build, re-check analysis.
• Clean up if genuinely messy, then done.
• Commit: "refactor: final cleanup"

=== POST-SHIP: WEB PREVIEW SMOKE GATE (CRITICAL — learned from deal-worthy recall 2026-05-22) ===

If the project has a deployable web target (Flutter web, Next.js, Vite, etc.) AND a configured preview channel (Firebase Hosting preview, Vercel preview, Cloudflare Pages preview), this gate runs BEFORE production promotion. It is not optional.

The 2026-04-25 deal-worthy hotfix cascade — Firebase init failure, missing /billing/success and /billing/cancel routes, broken Google Sign-In popup, and a dead-domain CTA — was 4 serial production hotfixes that a single preview-deploy smoke would have caught.

Procedure:

1. Build for web: flutter build web / npm run build / framework equivalent. Must succeed.
2. Deploy to a preview channel (NOT production):

• Firebase: firebase hosting:channel:deploy preview-{branch}-{ts} --only hosting
• Vercel: triggered automatically on PRs; if running locally, vercel --target preview
• Cloudflare Pages: wrangler pages deploy ... --branch preview-{ts}

1. Capture the preview URL.
2. Smoke-check the preview URL — at minimum hit every route the build added or modified plus all auth/billing/CTA routes. For each:

• HTTP 200 (or expected redirect to login).
• No console errors on initial load (Firebase init, Stripe init, analytics init).
• Auth provider sign-in popups open (don't have to complete sign-in, just confirm the popup launches without "redirect_uri_mismatch" or "popup_closed_by_user" instantly).
• External CTAs in the build resolve (no dead domains — curl -I the destinations).

1. If ANY smoke fails, STOP. Do not promote to production. Fix the failures and re-run the gate. Commit fixes as: "fix: preview smoke failures before production promote".
2. Only after smoke passes, promote: firebase hosting:clone preview-... live or merge to production branch.

This gate does NOT replace /qa or /bug-sweep — it is a thin pre-production smoke that catches the deploy-time class of failures (init, routes, popups, dead CTAs) that local builds don't surface.

=== POST-SHIP: DOCUMENTATION ===

After all iterations complete and validation passes:

• Run /readme to generate or update the project's README.md.

=== OUTPUT ===

One short summary:

Shipped

• What: [what you built]
• Pre-validation: [issues found/fixed, or "clean"]
• Status: [tests/build passing or not]
• Analysis: [issues found / issues fixed / any remaining]
• Server-side validation: [all wired / gaps found and fixed]
• Firestore rules: [all collections covered / gaps found and fixed]
• Documentation: [README.md generated/updated]
• Caveats: [any known issues, or "none"]

NEXT STEPS:

Recommended pipeline after /ship:

• "Run /qa to verify everything works end-to-end."
• "Run /e2e to generate automated end-to-end test coverage."
• "Run /iterate-review to harden with a focused review pass."
• "Run /ux to audit accessibility, design standards, and usability."
• "Run /polish for the full quality pipeline: /ux → /qa → /analyze."