@nockdev/awf 6.2.0

package/.agent/workflows/ap.md

@@ -0,0 +1,135 @@
+---
+description: "🔬 Full project audit (12-expert panel with conditional activation)"
+skills: { required: [security, audit-pro], contextual: [auto] }
+---
+
+# 🔬 /ap — Audit Pro
+
+> 12-Expert Panel Audit (5 Core + 7 Conditional) • 222 Checkpoints
+> 📊 ISO 25010 • CWE Top 25 • OWASP Top 10 • WCAG 2.2 • GDPR • SRE
+> 📁 Data: `skills/cross-cutting/audit-pro/data/` (HSA queryable)
+
+---
+
+## AUDIT FLOW
+
+1. **DISCOVERY** (Auto 30s) — Detect stack (`hsa_detect_stack`), count files, load audit-pro checklists via HSA (`hsa_get_context`), check audit history, **diff-aware**: if recent commits, focus on `git diff --name-only HEAD~5..HEAD`, **auto-activate** conditional experts based on detected project type
+2. **SCOPE CONTRACT** — Display scope options (from scoring.yaml) → ⛔ STOP wait for user selection (1-5). Show **previous audit score** if available for delta comparison. Show **active experts** based on detection
+3. **EXECUTE** — Run active Expert Panels sequentially, collect findings with evidence. Show **progress**: `[Panel 2/8] Architecture — Checkpoint 12/20`
+4. **SELF-REVIEW** — Agent re-reads findings, removes duplicates, verifies evidence accuracy, assigns confidence (1-10) per finding
+5. **REPORT** — Production Readiness Score (0-10), findings by P0/P1/P2/P3, **delta** vs previous audit (↑↓), save to `.domyh/audit_YYYY-MM-DD.md`
+6. **MEMORY PERSIST** — Update `memory/audit_summary.json`, log decisions, update `memory/state.json` scores
+
+---
+
+## EXPERT PANEL
+
+### Core (Always Active — 5)
+
+| ID           | EN Name        | VN Name | Seniority      | Focus            |
+| ------------ | -------------- | ------- | -------------- | ---------------- |
+| security     | David Chen     | Minh    | Principal 15yr | OWASP, CWE, Auth |
+| architecture | Sarah Kim      | Linh    | Staff 12yr     | SOLID, patterns  |
+| performance  | James Park     | Khoa    | Senior 10yr    | Latency, memory  |
+| quality      | Emma Wilson    | Hương   | Staff 12yr     | ISO 25010, tests |
+| devops       | Michael Torres | Đức     | Senior 10yr    | CI/CD, IaC       |
+
+### Conditional (Auto-detect — 7)
+
+| ID          | EN Name           | VN Name | Seniority          | Activates When                |
+| ----------- | ----------------- | ------- | ------------------ | ----------------------------- |
+| ux          | Lisa Wang         | Thảo    | Senior 8yr         | UI files/deps detected        |
+| data        | Robert Martinez   | Tuấn    | Principal 15yr     | DB/migration files detected   |
+| compliance  | Jennifer Anderson | Hà      | Distinguished 18yr | Regulated industry indicators |
+| product     | Daniel Lee        | Lan     | Tech Lead 10yr     | `scope_full` or user request  |
+| reliability | William Brown     | Phong   | Staff 12yr         | Production/deploy files       |
+| cloud       | Alexander White   | Bảo     | Senior 10yr        | IaC/Terraform/K8s files       |
+| ai_safety   | Dr. Sophia Nguyen | Vy      | Distinguished 20yr | AI/ML/LLM code/deps           |
+
+---
+
+## DIFF-AWARE MODE
+
+```yaml
+# When recent changes detected:
+diff_aware:
+  trigger: "git log -1 --since='7 days ago'"
+  scope: "git diff --name-only HEAD~5..HEAD"
+  behavior: "Focus audit on changed files + their dependents"
+  full_scan: "Still run full scan but prioritize changed files first"
+```
+
+---
+
+## TOOL INTEGRATION (Optional)
+
+> If tools available in project, auto-run BEFORE AI audit for higher accuracy:
+
+```yaml
+pre_audit_tools:
+  javascript:
+    - "npx eslint --format json src/" # Lint issues
+    - "npm audit --json" # Dependency vulns
+    - "npx semgrep --config auto --json" # SAST security
+  python:
+    - "ruff check --output-format json" # Lint
+    - "pip-audit --format json" # Dependency vulns
+    - "bandit -r src/ -f json" # Security
+  go:
+    - "go vet ./..." # Static analysis
+    - "govulncheck ./..." # Vulnerability check
+    - "golangci-lint run --out-format json" # Lint
+  general:
+    - "trivy fs --format json ." # Container/FS scan
+
+
+# Tool results are fed as additional context to Expert Panels
+# This achieves ~95% accuracy (vs ~78% AI-only)
+```
+
+---
+
+## DATA LOADING
+
+```yaml
+checklists: skills/cross-cutting/audit-pro/data/checklists.yaml # 12 experts, 222 checkpoints
+scoring: skills/cross-cutting/audit-pro/data/scoring.yaml # 6 weight profiles, grades, scope options
+```
+
+---
+
+## FINDING FORMAT
+
+```
+**[P0]** 🔒 Expert `file:line` (Confidence: 9/10)
+**Issue:** description  **Evidence:** code snippet  **Impact:** risk  **Fix:** suggested fix
+```
+
+---
+
+## REPORT FORMAT
+
+```
+📊 DOMYH AUDIT — [project] — [date] — Score: X.X/10 (↑0.3 from last)
+| Expert | Score | Issues | Δ vs Last |
+👥 Active Experts: 8/12 (Security, Architecture, Performance, Quality, DevOps, UX, Data, Reliability)
+📁 Full: .domyh/audit_YYYY-MM-DD.md
+⏱️ Duration: Xm Ys | Files: N | Changed: M
+```
+
+---
+
+## SUB-COMMANDS
+
+| Command              | Description                           |
+| -------------------- | ------------------------------------- |
+| `/ap`                | Full audit (all active experts)       |
+| `/ap quick`          | Quick audit (Security + Architecture) |
+| `/ap security`       | Security focus only                   |
+| `/ap performance`    | Performance focus only                |
+| `/ap expert [name]`  | Single expert audit                   |
+| `/ap --scope [path]` | Limit scope                           |
+| `/ap --diff`         | Diff-aware (changed files only)       |
+| `/ap --compare`      | Compare with previous audit           |
+| `/ap --experts`      | Show active/inactive expert status    |
+| `/ap --force [name]` | Force-include a conditional expert    |

package/.agent/workflows/code.md

@@ -0,0 +1,130 @@
+---
+description: "💻 Write production-ready code, fix/improve existing projects, with proper error handling, types, and documentation"
+skills: { required: [], contextual: [auto] }
+---
+
+# 💻 /code — Code Pro
+
+> Intelligent code generation & project improvement with language-specific patterns
+> 📚 30+ Languages • Auto Test Loop • Fix/Improve Mode • Self-Review
+
+---
+
+## CODE FLOW
+
+1. **DETECT** (Auto) — Parse intent (feature/bugfix/refactor), detect stack via HSA (`hsa_detect_stack`), load language skill via HSA (`hsa_get_context`)
+2. **PLAN** — Break down into steps, identify dependencies → `hsa_prefetch` planned files → ⛔ STOP if major change (>50 lines). Show step count: `Plan: 4 steps, ~85 lines`
+3. **EXECUTE** — Write code following skill patterns, error handling + types, add tests (auto test loop). Show progress: `[Step 2/4] Creating auth middleware...`
+4. **VERIFY** — Run tests → Fix → Repeat (max 3), lint check. Agent re-reads ALL generated code:
+   - [ ] Matches original intent?
+   - [ ] Edge cases handled?
+   - [ ] No hardcoded values or magic numbers?
+   - [ ] Error messages helpful and descriptive?
+   - [ ] Names follow language conventions?
+   - [ ] No security anti-patterns (hardcoded secrets, raw SQL)?
+   - → If issues found: fix silently before output
+5. **SYNC** — `hsa_check_changes` to update index, output summary of changes, confidence score (1-10), next steps. Persist key decisions to `memory/state.json`
+
+---
+
+## SUB-COMMANDS
+
+| Command                        | Description             | Mode     |
+| ------------------------------ | ----------------------- | -------- |
+| `/code [task]`                 | Generate code for task  | Create   |
+| `/code fix [issue]`            | Fix existing issue      | Fix      |
+| `/code improve [area]`         | Improve existing code   | Refactor |
+| `/code add [feature]`          | Add feature to existing | Update   |
+| `/code test [feature]`         | Generate tests only     | Test     |
+| `/code secure [feature]`       | Generate secure code    | Create   |
+| `/code quality analyze [path]` | Static analysis         | Analyze  |
+
+---
+
+## FIX/IMPROVE MODE
+
+1. **DETECT** stack (auto) → 2. **ANALYZE** issue (scan files) → 3. **PLAN** fix (scope, confirm) → 4. **EXECUTE** fix → 5. **VERIFY** (tests, lint) → 6. **SELF-REVIEW** → 7. **SUMMARY**
+
+### Priority Matrix
+
+| Priority | Description                  | SLA          | Action               |
+| -------- | ---------------------------- | ------------ | -------------------- |
+| **P0**   | Critical security/breaking   | Immediate    | Must fix now         |
+| **P1**   | Affects core functionality   | Same session | Fix before next task |
+| **P2**   | Code quality/maintainability | This sprint  | Schedule fix         |
+| **P3**   | Nice to have improvements    | Backlog      | Track                |
+
+---
+
+## LANGUAGE PATTERNS (Quick Reference)
+
+> Full patterns loaded via language skills (HSA). Each skill has: patterns, style, error handling, structure, tests.
+
+| Language   | Patterns               | Style           | Error Handling           | Tests           |
+| ---------- | ---------------------- | --------------- | ------------------------ | --------------- |
+| Go         | Interfaces, Channels   | gofmt           | error wrapping           | go test         |
+| Python     | Type hints, Decorators | black, ruff     | try-except               | pytest          |
+| TypeScript | Strict types, Generics | prettier+eslint | try-catch + custom Error | jest, vitest    |
+| Rust       | Ownership, Traits      | rustfmt         | Result<T,E>              | cargo test      |
+| Java       | Spring DI, Streams     | google-java     | try-catch                | JUnit 5         |
+| C#         | .NET Core, LINQ        | dotnet format   | try-catch, nullable      | xUnit           |
+| PHP        | Laravel patterns       | PHP-CS-Fixer    | try-catch                | PHPUnit         |
+| Ruby       | Rails conventions      | rubocop         | begin-rescue             | RSpec           |
+| Swift      | Protocols, Actors      | swift-format    | do-try-catch             | XCTest          |
+| Kotlin     | Coroutines, DSL        | ktlint          | runCatching              | kotest          |
+| C++        | RAII, Smart ptrs       | clang-format    | exceptions               | GoogleTest      |
+| React      | Hooks, Context         | prettier        | ErrorBoundary            | RTL             |
+| Next.js    | App Router, RSC        | prettier        | error.tsx                | Jest+Playwright |
+| Vue        | Composition API        | eslint-vue      | onErrorCaptured          | Vitest          |
+| Angular    | Signals, RxJS          | Angular Guide   | ErrorHandler             | Jasmine         |
+
+> Mobile: React Native, Flutter, Swift(iOS), Kotlin(Android)
+> Functional: Elixir, Haskell, Scala, OCaml
+> Data/ML: Python+NumPy, R, Julia
+
+---
+
+## CODE QUALITY CONTRACT
+
+```yaml
+every_change:
+  - Error handling at every boundary
+  - Input validation on public APIs
+  - Types/interfaces defined
+  - Named exports preferred
+  - Constants over magic numbers
+
+test_loop:
+  max_iterations: 3
+  strategy: "Write test → Run → Fix → Repeat"
+  coverage_target: "> 70% (new code)"
+
+self_review:
+  enabled: true
+  trigger: "after_verify, before_output"
+  checklist: 7 # items checked above
+```
+
+---
+
+## SECURITY-FIRST
+
+| Category | Default Behavior                         |
+| -------- | ---------------------------------------- |
+| Input    | Validate all user input, sanitize        |
+| Output   | Context-aware encoding, XSS prevention   |
+| Errors   | Non-revealing errors, structured logging |
+| Secrets  | Environment variables, no hardcoded      |
+| Queries  | Parameterized only (no raw SQL)          |
+| Auth     | RBAC checks, secure sessions             |
+
+---
+
+## AI QUALITY GATES
+
+| Layer | Name            | What                                       |
+| ----- | --------------- | ------------------------------------------ |
+| 1     | AI Auto-Fix     | Linting, formatting, naming, comments      |
+| 2     | Static Analysis | Snyk, SonarQube, ESLint; cyclomatic < 10   |
+| 3     | Self-Review     | Agent self-critique before delivery        |
+| 4     | Human Review    | Complex logic, security, infra, >100 lines |

package/.agent/workflows/debug.md

@@ -0,0 +1,230 @@
+---
+description: "🐛 Systematic debugging: reproduce → isolate → analyze → fix → verify"
+skills: { required: [], contextual: [auto] }
+---
+
+# 🐛 /debug — Debug Pro
+
+> AI-Powered Debugging with Observability & Tracing
+> 📚 30+ Languages • Root Cause Analysis • Hypothesis Testing • Failure Repository
+
+---
+
+## DEBUG FLOW
+
+1. **CAPTURE** (Auto) — Auto-detect language via HSA (`hsa_detect_stack`), load skill via HSA (`hsa_get_context`), parse stack trace, identify affected files. Show: `[Step 1/7] Capturing error context...`
+2. **TIMELINE** — Reconstruct event timeline: `git log --oneline -10`, check recent changes to affected files, correlate with error timestamps. Answer: "When did this start?"
+3. **REPRODUCE** — Create minimal reproduction, confirm error occurs consistently → ⛔ STOP if cannot reproduce: ask user 5 questions
+4. **ISOLATE** — Binary search / git bisect, add trace logging, narrow to exact location
+5. **HYPOTHESIZE** — Form 2-3 hypotheses based on evidence:
+   ```
+   H1: [hypothesis] — Evidence: [what supports it] — Test: [how to verify]
+   H2: [hypothesis] — Evidence: [what supports it] — Test: [how to verify]
+   → Test each hypothesis systematically → Confirm/reject
+   ```
+6. **ANALYZE** — 5 Whys on CONFIRMED root cause (not assumptions)
+7. **FIX** — Create FAILING test first, implement single fix → if 3+ fixes fail → question architecture. Add **prevention guard** to block similar bugs
+8. **VERIFY** — Run reproduction steps, run full test suite, show before/after evidence. Persist pattern to failure repository if novel
+
+---
+
+## DEBUG MINDSET
+
+```
+❌ NEVER assume these:
+- "I know what the bug is"         → Always reproduce first
+- "I'll be quick"                  → Quick fixes create more bugs
+- "It worked before"               → Something changed — check timeline
+- "It must be a library bug"       → It's almost never the library
+- "One fix should do it"           → Form hypotheses first, test each
+```
+
+---
+
+## TIMELINE RECONSTRUCTION
+
+```yaml
+# Before diving into code, reconstruct the timeline:
+timeline:
+  sources:
+    - "git log --oneline --since='7 days ago' -- [affected_files]"
+    - "Error logs / console output (timestamps)"
+    - "Recent dependency updates (package-lock.json diff)"
+    - "Environment changes (.env, config files)"
+  output: |
+    [timestamp] Event 1: Last known working state
+    [timestamp] Event 2: Change X committed
+    [timestamp] Event 3: Error first reported
+    → Most likely cause: Event between 2 and 3
+```
+
+---
+
+## HYPOTHESIS TESTING
+
+```yaml
+# Instead of jumping to fix, test hypotheses:
+process:
+  1_form: "Based on evidence, create 2-3 hypotheses"
+  2_rank: "Rank by likelihood (most evidence first)"
+  3_test: "Design quick test for top hypothesis"
+  4_confirm: "If confirmed → proceed to 5 Whys"
+  5_reject: "If rejected → test next hypothesis"
+  6_stuck: "If all rejected → expand search scope, ask user"
+
+example:
+  H1: "Null user from DB query (no validation)"
+    evidence: "TypeError at user.id, findUser() called without check"
+    test: "Add console.log before line 42, check user value"
+    result: "CONFIRMED — user is undefined when ID not found"
+```
+
+---
+
+## STACK TRACE PARSERS
+
+| Language   | Error Pattern                                | Debugger          |
+| ---------- | -------------------------------------------- | ----------------- |
+| Go         | `goroutine \d+ [running]:, panic:`           | dlv               |
+| Python     | `Traceback (most recent call last):`         | pdb, debugpy      |
+| TypeScript | `at Object.<anonymous> (.ts:\d+)`            | --inspect-brk     |
+| Rust       | `thread 'main' panicked at`                  | rust-gdb          |
+| Java       | `at com.package.Class.method(File.java:\d+)` | jdb               |
+| C++        | `Segmentation fault\|terminate called`       | gdb, lldb         |
+| C#         | `Unhandled exception.*at .*\.cs:line`        | dotnet-dump       |
+| PHP        | `Fatal error:.*in .*.php on line`            | xdebug            |
+| Ruby       | `from .*.rb:\d+:in`                          | byebug            |
+| Swift      | `Fatal error:.*file .*\.swift, line`         | lldb              |
+| Kotlin     | `at .*\.kt:\d+`                              | IntelliJ debugger |
+
+> Full parsers for 30+ languages via HSA: each skill's META.yaml contains `triggers.file_patterns`.
+
+---
+
+## ⛔ IF CANNOT REPRODUCE
+
+```
+Need more information:
+1. Exact error message (copy-paste)?
+2. Steps to trigger the error?
+3. Environment details (OS, versions)?
+4. Recent changes (git log)?
+5. Logs from when error occurred?
+→ If intermittent: suspect race condition, timing, or environment-specific issue
+```
+
+---
+
+## GIT BISECT
+
+```bash
+git bisect start
+git bisect bad HEAD
+git bisect good <last_working_commit>
+# Agent can automate: git bisect run <test_script>
+```
+
+## QUICK DEBUG LOGGING
+
+| Language   | Snippet                                      |
+| ---------- | -------------------------------------------- |
+| Go         | `log.Printf("[DEBUG] %s: %+v", "var", val)`  |
+| Rust       | `dbg!(&variable);`                           |
+| Python     | `logging.debug(f"[DEBUG] {variable=}")`      |
+| TypeScript | `console.log('[DEBUG]', { variable });`      |
+| Java       | `System.out.println("[DEBUG] " + variable);` |
+
+---
+
+## ROOT CAUSE ANALYSIS (5 Whys)
+
+```
+Why 1: [immediate cause]
+Why 2: [why that happened]
+Why 3: [deeper reason]
+Why 4: [systemic issue]
+Why 5: [root cause]
+
+🎯 ROOT CAUSE: [one-line summary]
+🛡️ PREVENTION: [guard to prevent recurrence]
+```
+
+## BUG CATEGORIES
+
+| Category       | Pattern                      | Common Fix                       |
+| -------------- | ---------------------------- | -------------------------------- |
+| Null Reference | `null\|undefined\|None\|nil` | Null checks, optional chaining   |
+| Type Error     | `TypeError\|cannot convert`  | Validate types, add guards       |
+| Async Race     | `race condition\|deadlock`   | Add locks, atomic ops            |
+| Memory Leak    | `out of memory\|heap`        | Cleanup, weak references         |
+| Network        | `timeout\|ETIMEDOUT`         | Retry logic, increase timeout    |
+| Permission     | `EACCES\|unauthorized`       | Check file/API permissions       |
+| State          | `stale data\|inconsistent`   | Cache invalidation, transactions |
+
+---
+
+## OBSERVABILITY INTEGRATION
+
+```yaml
+# If observability tools available, use them:
+tools:
+  sentry: "Check Sentry for error frequency, affected users, breadcrumbs"
+  datadog: "Query DataDog APM for traces, latency spikes"
+  grafana: "Check Grafana dashboards for metric anomalies"
+  cloudwatch: "Search CloudWatch logs for error patterns"
+
+# Ask user: "Do you have any monitoring/APM tools? (Sentry, DataDog, etc.)"
+```
+
+---
+
+## DEFENSE-IN-DEPTH
+
+> Validate at EVERY layer data passes through: "We made the bug impossible"
+
+| Layer                    | Purpose                               | Example                                 |
+| ------------------------ | ------------------------------------- | --------------------------------------- |
+| 1. Entry Point           | Reject invalid input at API boundary  | `if (!dir) throw Error("dir required")` |
+| 2. Business Logic        | Ensure data makes sense for operation | Validate IDs, check states              |
+| 3. Environment Guard     | Prevent danger in specific contexts   | Refuse ops outside temp in tests        |
+| 4. Debug Instrumentation | Capture context for forensics         | Stack traces, debug logs                |
+
+---
+
+## FAILURE REPOSITORY
+
+```yaml
+failure_patterns:
+  - id: "cache-invalidation"
+    symptoms: ["stale data", "deleted item still shows"]
+    fix: "Add cache.invalidate() after mutations"
+    prevention: "Always invalidate cache in write operations"
+  - id: "n+1-query"
+    symptoms: ["slow API", "many DB queries"]
+    fix: "Use eager loading / batch queries"
+    prevention: "Use query analyzer / ORM eager loading"
+  - id: "race-condition"
+    symptoms: ["intermittent failures", "data corruption"]
+    fix: "Add mutex/transaction"
+    prevention: "Always use transactions for multi-step writes"
+  - id: "off-by-one"
+    symptoms: ["missing last item", "index out of bounds"]
+    fix: "Check loop boundaries, use length-1"
+    prevention: "Prefer forEach/map over manual indexing"
+```
+
+---
+
+## SUB-COMMANDS
+
+| Command                  | Description                |
+| ------------------------ | -------------------------- |
+| `/debug [error]`         | Full debug flow            |
+| `/debug ai [error]`      | AI-powered analysis        |
+| `/debug similar [error]` | Search failure repository  |
+| `/debug save [pattern]`  | Save to repository         |
+| `/debug --trace`         | Add verbose logging        |
+| `/debug --bisect`        | Git bisect helper          |
+| `/debug --profile`       | Performance profiling      |
+| `/debug --memory`        | Memory leak detection      |
+| `/debug --timeline`      | Reconstruct event timeline |

package/.agent/workflows/deploy.md

@@ -0,0 +1,192 @@
+---
+description: "🚀 Deploy to production with pre-checks, rollback plan, and post-verification"
+skills: { required: [ci-cd], contextual: [docker, kubernetes, aws] }
+---
+
+# 🚀 /deploy — Deploy Pro
+
+> Zero-Downtime Deployments with Rollback Safety
+> 📚 16 Platforms • Progressive Delivery • GitOps • Post-Deploy Monitoring
+
+---
+
+## DEPLOYMENT FLOW
+
+1. **PRE-FLIGHT** (Auto) — Detect stack via HSA (`hsa_detect_stack`), load deploy context (`hsa_get_context`). Run pre-flight checklist:
+   - [ ] Tests pass (`npm test` / `go test ./...`)
+   - [ ] Security scan (no P0 vulnerabilities)
+   - [ ] Dependencies up to date (no critical outdated)
+   - [ ] Environment variables configured
+   - [ ] Database migrations pending? → Run first
+   - [ ] Feature flags configured?
+   - [ ] Rollback plan documented
+   - [ ] Team notified? (Slack/Discord/Email)
+         → ⛔ STOP if any critical check fails. Show: `[Pre-flight 6/8] ✅`
+2. **BUILD** — Create production build, generate version tag, create artifacts
+3. **DEPLOY** — Execute strategy (rolling/canary/blue-green/recreate), health checks at each stage. Show: `[Deploy] Canary 5% → monitoring...`
+4. **VERIFY** — Smoke tests, health endpoints, error rate <0.1%, latency p99 <500ms, CPU <80%
+5. **MONITOR** — Watch metrics for 15 minutes post-deploy:
+   ```
+   ⏱️ Monitoring window: 15min remaining...
+   📊 Error rate: 0.02% ✅ | p99: 280ms ✅ | CPU: 45% ✅
+   → Auto-rollback if: error_rate > 0.5% OR p99 > 1000ms OR CPU > 90%
+   ```
+6. **FINALIZE** — Tag release in git, update changelog, document rollback point, release deploy lock
+
+---
+
+## PRE-FLIGHT CHECKLIST
+
+```yaml
+pre_flight:
+  required:
+    - test_pass: "All tests must pass"
+    - no_p0_vulns: "No P0 security vulnerabilities"
+    - env_vars: "All required env vars set"
+    - rollback_plan: "Rollback steps documented"
+  recommended:
+    - db_migrations: "Run pending migrations first"
+    - feature_flags: "Toggle flags for new features"
+    - team_notify: "Notify team in communication channel"
+    - staging_tested: "Verified in staging environment"
+    - changelog: "CHANGELOG.md updated"
+
+  on_fail:
+    required: "⛔ STOP — Cannot deploy"
+    recommended: "⚠️ WARNING — Proceed with caution?"
+```
+
+---
+
+## DEPLOY LOCK
+
+```yaml
+# Prevent concurrent deployments
+deploy_lock:
+  file: ".deploy.lock"
+  content: |
+    deployer: [user/agent]
+    started: [ISO timestamp]
+    environment: [production/staging]
+    version: [tag]
+  behavior:
+    on_lock_exists: "⛔ STOP — Another deployment in progress"
+    on_complete: "Remove lock file"
+    on_timeout: "Auto-release after 30 minutes"
+```
+
+---
+
+## PLATFORM REGISTRY
+
+### Cloud / PaaS
+
+```yaml
+# platform: build | deploy | rollback | health
+vercel: npm run build | vercel --prod | vercel rollback | automatic
+netlify: npm run build | netlify deploy --prod | netlify rollback | deploy previews
+cloudflare: npm run build | wrangler pages deploy | dashboard revert | automatic
+gh_pages: npm run build | gh-pages -d dist | git revert + deploy | manual
+railway: auto from git | railway up | dashboard | automatic
+fly_io: fly deploy | fly deploy --strategy rolling | fly releases rollback | fly status
+render: auto from git | push to branch | dashboard | health path
+heroku: auto buildpacks | git push heroku main | heroku releases:rollback | metrics
+```
+
+### Containers / Cloud
+
+```yaml
+docker: docker build -t app . | docker-compose up -d | tag previous | HEALTHCHECK
+kubernetes: docker build + apply | kubectl apply -f k8s/ | kubectl rollout undo | liveness/readiness
+aws_ecs: docker build + ecr push | aws ecs update-service | --task-definition prev | ALB health
+aws_lambda: sam build | sam deploy --guided | --s3-key previous.zip | CloudWatch
+gcp_run: gcloud builds submit | gcloud run deploy | --to-revisions=REV=100 | container health
+azure: az acr build | az webapp deployment | slot swap | health path
+```
+
+### VPS / Bare Metal
+
+```yaml
+vps_ssh: local/remote build | scp + ssh | keep prev release folder | curl health
+pm2: npm run build | pm2 deploy production | pm2 deploy revert 1 | pm2 monit
+systemd: build locally | scp + systemctl restart | keep prev binary | systemctl status
+```
+
+### Build Commands by Stack
+
+```yaml
+# Frontend: react/vue/svelte/angular → npm run build | nextjs → next build
+# Backend: go build -o app -ldflags='-s -w' | cargo build --release | mvn package | dotnet publish -c Release
+# Mobile: xcodebuild archive | ./gradlew assembleRelease | flutter build apk --release
+```
+
+---
+
+## DEPLOYMENT STRATEGIES
+
+| Strategy       | Rollback              | Risk            | Cost     | Use When                 |
+| -------------- | --------------------- | --------------- | -------- | ------------------------ |
+| **Blue-Green** | Instant (switch back) | Low             | 2x infra | Zero-downtime required   |
+| **Canary**     | Route back to stable  | Very Low        | Minimal  | High-risk, metrics-based |
+| **Rolling**    | Rollout undo          | Medium          | Minimal  | Standard K8s deploy      |
+| **Recreate**   | Deploy prev version   | High (downtime) | Lowest   | Dev/staging only         |
+
+### Canary Stages: 1% → 5% → 25% → 100% (10min/stage, auto-rollback on threshold breach)
+
+---
+
+## POST-DEPLOY MONITORING
+
+```yaml
+monitoring:
+  duration: 15min # Watch window after deploy
+  interval: 30s # Check frequency
+
+  thresholds:
+    error_rate: 0.5% # Auto-rollback if exceeded
+    p99_latency: 1000ms # Auto-rollback if exceeded
+    cpu_usage: 90% # Warning, manual decision
+    memory: 85% # Warning, manual decision
+
+  actions:
+    breach_critical: "AUTO-ROLLBACK + notify team"
+    breach_warning: "ALERT team, continue monitoring"
+    all_clear: "✅ Deploy stable, proceed to finalize"
+```
+
+---
+
+## HEALTH CHECKS
+
+```yaml
+checks:
+  http: ["/health → 200", "/api/status → 200 {status:ok}", "/ → 200"]
+  smoke:
+    [
+      Homepage loads,
+      API returns data,
+      Auth flow works,
+      Critical path functional,
+    ]
+  metrics: [error_rate <0.1% (5m), p99_latency <500ms (5m), cpu <80%]
+```
+
+---
+
+## SUB-COMMANDS
+
+| Command                          | Description           |
+| -------------------------------- | --------------------- |
+| `/deploy`                        | Deploy to production  |
+| `/deploy staging`                | Deploy to staging     |
+| `/deploy --dry`                  | Dry run (preview)     |
+| `/deploy --force`                | Skip pre-checks ⚠️    |
+| `/deploy --canary`               | Force canary strategy |
+| `/deploy gitops sync`            | GitOps sync           |
+| `/deploy progressive [strategy]` | Progressive delivery  |
+| `/deploy monitor`                | AI deployment monitor |
+| `/deploy --checklist`            | Show pre-flight only  |
+| `/deploy readiness`              | SRE readiness check   |
+| `/deploy signoff`                | Stakeholder sign-off  |
+| `/deploy validate-prompts`       | AI prompt validation  |
+| `/rollback`                      | Rollback to previous  |