@agentskillkit/agent-skills 1.0.1 → 3.2.1

package/.agent/workflows/autopilot.md

@@ -1,8 +1,11 @@
+---
+description: Autonomous agent-patterns coordination. 3+ specialists working in parallel with verification.
+chain: build-web-app
 ---
-description: Autonomous multi-agent coordination. 3+ specialists working in parallel with verification.
----
 
-# /autopilot - Multi-Agent Command Center
+// turbo-all
+
+# /autopilot - agent-patterns Command Center
 
 $ARGUMENTS
 
@@ -14,23 +17,23 @@ Coordinate 3+ specialist agents for complex tasks. **Like having a senior engine
 
 ---
 
-## 🔴 CRITICAL: Minimum 3 Agents
+## 🔴 CRITICAL: Minimum 3 Agents
 
 > **AUTOPILOT = MINIMUM 3 DIFFERENT SPECIALISTS**
-> 
-> If fewer than 3 agents → NOT autopilot, just delegation.
-> Single agent work → Use direct command instead.
+>
+> If fewer than 3 agents → NOT autopilot, just delegation.
+> Single agent work → Use direct command instead.
 
 ---
 
 ## Pre-Flight: Mode Check
 
-| Current Mode | Task Type | Action |
-|--------------|-----------|--------|
-| **plan** | Any | ✅ Proceed with planning-first approach |
-| **edit** | Simple | ✅ Proceed directly |
-| **edit** | Complex | ⚠️ Ask: "This requires planning. Switch to plan mode?" |
-| **ask** | Any | ⚠️ Ask: "Ready to orchestrate. Switch to edit or plan mode?" |
+| Current Mode | Task Type | Action                                                       |
+| ------------ | --------- | ------------------------------------------------------------ |
+| **plan**     | Any       | ✅ Proceed with planning-first approach                      |
+| **edit**     | Simple    | ✅ Proceed directly                                          |
+| **edit**     | Complex   | ⚠️ Ask: "This requires planning. Switch to plan mode?"       |
+| **ask**      | Any       | ⚠️ Ask: "Ready to orchestrate. Switch to edit or plan mode?" |
 
 ---
 
@@ -39,67 +42,302 @@ Coordinate 3+ specialist agents for complex tasks. **Like having a senior engine
 Before selecting agents, identify ALL domains:
 
 ```
-□ Security     → security-auditor
-□ Backend/API  → backend-specialist
-□ Frontend/UI  → frontend-specialist
-□ Database     → database-architect
-□ Testing      → test-engineer
-□ DevOps       → devops-engineer
-□ Mobile       → mobile-developer
-□ Performance  → performance-optimizer
-□ SEO          → seo-specialist
-□ Planning     → project-planner
-□ Debug        → debugger
+□ Security     → security-auditor
+□ Backend/API  → backend-specialist
+□ Frontend/UI  → frontend-specialist
+□ Database     → database-architect
+□ Testing      → test-engineer
+□ DevOps       → devops-engineer
+□ Mobile       → mobile-developer
+□ Performance  → performance-optimizer
+□ SEO          → seo-specialist
+□ Planning     → project-planner
+□ Debug        → debugger
 ```
 
 ---
 
 ## Agent Selection Matrix
 
-| Task Type | Required Agents | Verification |
-|-----------|-----------------|--------------| 
-| **Web App** | frontend, backend, test | lint + security |
-| **API** | backend, security, test | security scan |
-| **UI/Design** | frontend, seo, performance | lighthouse |
-| **Database** | database, backend, security | schema check |
-| **Full Stack** | planner, frontend, backend, devops | all |
-| **Debug** | debugger, explorer, test | reproduction test |
-| **Security** | security-auditor, penetration-tester, devops | all scans |
+| Task Type      | Required Agents                              | Verification      |
+| -------------- | -------------------------------------------- | ----------------- |
+| **Web App**    | frontend, backend, test                      | lint + security   |
+| **API**        | backend, security, test                      | security scan     |
+| **UI/Design**  | frontend, seo, performance                   | lighthouse        |
+| **Database**   | database, backend, security                  | schema check      |
+| **Full Stack** | planner, frontend, backend, devops           | all               |
+| **Debug**      | debugger, explorer, test                     | reproduction test |
+| **Security**   | security-auditor, penetration-tester, devops | all scans         |
+
+---
+
+## Available Agents (22 total)
+
+| Agent                   | Domain    | Use When                |
+| ----------------------- | --------- | ----------------------- |
+| `project-planner`       | Planning  | Task breakdown, PLAN.md |
+| `explorer-agent`        | Discovery | Codebase mapping        |
+| `frontend-specialist`   | UI/UX     | React, Vue, CSS, HTML   |
+| `backend-specialist`    | Server    | API, Node.js, Python    |
+| `database-architect`    | Data      | SQL, NoSQL, Schema      |
+| `security-auditor`      | Security  | Vulnerabilities, Auth   |
+| `test-engineer`         | Testing   | Unit, E2E, Coverage     |
+| `devops-engineer`       | Ops       | CI/CD, Docker, Deploy   |
+| `mobile-developer`      | Mobile    | React Native, Flutter   |
+| `performance-optimizer` | Speed     | Lighthouse, Profiling   |
+| `seo-specialist`        | SEO       | Meta, Schema, Rankings  |
+| `documentation-writer`  | Docs      | README, API docs        |
+| `debugger`              | Debug     | Error analysis          |
+
+### Meta-Agents (Runtime Control)
+
+| Agent         | Role       | When Invoked                    |
+| ------------- | ---------- | ------------------------------- |
+| `orchestrator`| Runtime    | Controls execution, retry, parallelism |
+| `assessor`    | Risk       | Before risky operations         |
+| `recovery`    | Safety     | Save state, rollback on failure |
+| `critic`      | Arbitration| Resolve QA vs Execution conflicts |
+| `learner`     | Learning   | After failures, extract lessons |
+
+## 🤖 Meta-Agents Integration
+
+```
+Flow:
+orchestrator.init() → assessor.evaluate(plan)
+       ↓
+recovery.save() → execute phases in parallel
+       ↓
+conflict? → critic.arbitrate()
+       ↓
+failure? → recovery.restore() → learner.log()
+       ↓
+success → learner.log(patterns)
+```
+
+---
+
+## 🔒 MANDATORY: Meta-Agent Hooks
+
+> **ENFORCEMENT:** These hooks are REQUIRED for every autopilot execution.
+> Failure to invoke meta-agents = incomplete autopilot workflow.
+
+### Hook 1: Pre-Flight Risk Assessment (assessor)
+
+**When:** BEFORE Phase 2 execution starts  
+**Agent:** `assessor` (impact-assessor)
+
+```markdown
+## Pre-Flight Assessment Protocol
+
+1. Invoke assessor with plan summary
+2. Receive risk evaluation:
+   - Risk Level: LOW / MEDIUM / HIGH / CRITICAL
+   - Impact Scope: [files, services affected]
+   - Mitigation Recommendations
+3. If CRITICAL → Require explicit user confirmation
+4. If HIGH → Log warning, proceed with extra checkpoints
+5. If LOW/MEDIUM → Proceed normally
+```
+
+**Invocation Template:**
+```
+Use the assessor agent to evaluate risk:
+- Task: [task summary]
+- Affected domains: [list]
+- Estimated scope: [files/services]
+```
+
+---
+
+### Hook 2: State Preservation (recovery)
+
+**When:** BEFORE any risky phase starts  
+**Agent:** `recovery` (recovery-agent)
+
+```markdown
+## Checkpoint Protocol
+
+1. BEFORE execution: recovery.saveState(affectedFiles)
+2. Store checkpoint ID for rollback
+3. IF failure occurs: recovery.restore(checkpointId)
+4. AFTER success: cleanup old checkpoints (keep last 3)
+```
+
+**Risky Operations (REQUIRE checkpoint):**
+- Database migrations
+- Multi-file refactoring
+- Configuration changes
+- Dependency updates
+- Build/deploy operations
+
+**Invocation Template:**
+```
+Use the recovery agent to save state:
+- Operation: [description]
+- Affected files: [list]
+- Reason: [why checkpoint needed]
+```
+
+---
+
+### Hook 3: Conflict Resolution (critic)
+
+**When:** Domain agents have conflicting recommendations  
+**Agent:** `critic` (critic-judge)
+
+```markdown
+## Conflict Resolution Protocol
+
+1. Detect conflict between agents (e.g., QA vs speed)
+2. Invoke critic with both positions
+3. Critic applies decision hierarchy:
+   - Safety > Recoverability > Correctness > Cleanliness > Convenience
+4. Follow critic's verdict
+```
+
+**Common Conflicts:**
+- `test-engineer` vs `devops-engineer` (coverage vs deploy speed)
+- `security-auditor` vs `backend-specialist` (security vs functionality)
+- `performance-optimizer` vs `frontend-specialist` (speed vs UX)
+
+**Invocation Template:**
+```
+Use the critic agent to resolve conflict:
+- Party A: [agent] recommends [action] because [reason]
+- Party B: [agent] recommends [action] because [reason]
+- Context: [relevant background]
+```
+
+---
+
+### Hook 4: Learning & Improvement (learner)
+
+**When:** After EVERY autopilot execution (success OR failure)  
+**Agent:** `learner` (learning-agent)
+
+```markdown
+## Learning Protocol
+
+1. ON FAILURE:
+   - learner.logFailure(error, context, rootCause)
+   - Add to lessons-learned.yaml
+   - Tag pattern for future prevention
+
+2. ON SUCCESS:
+   - learner.logSuccess(patterns, optimizations)
+   - Record what worked well
+   - Update best practices
+
+3. observability INTEGRATION:
+   - Log execution observability from AutopilotMetrics
+   - Compare to baseline
+   - Track improvement trends
+```
+
+**Invocation Template (Failure):**
+```
+Use the learner agent to log failure:
+- Error: [what happened]
+- Phase: [which phase failed]
+- Root cause: [if known]
+- How it was resolved: [fix applied or user intervention]
+```
+
+**Invocation Template (Success):**
+```
+Use the learner agent to log success patterns:
+- Task: [task summary]
+- Agents used: [list]
+- Key optimizations: [what made it efficient]
+- Duration: [total time]
+```
 
 ---
 
-## Available Agents (17 total)
+### Meta-Agent Enforcement Checklist
+
+Before completing autopilot, verify:
 
-| Agent | Domain | Use When |
-|-------|--------|----------|
-| `project-planner` | Planning | Task breakdown, PLAN.md |
-| `explorer-agent` | Discovery | Codebase mapping |
-| `frontend-specialist` | UI/UX | React, Vue, CSS, HTML |
-| `backend-specialist` | Server | API, Node.js, Python |
-| `database-architect` | Data | SQL, NoSQL, Schema |
-| `security-auditor` | Security | Vulnerabilities, Auth |
-| `test-engineer` | Testing | Unit, E2E, Coverage |
-| `devops-engineer` | Ops | CI/CD, Docker, Deploy |
-| `mobile-developer` | Mobile | React Native, Flutter |
-| `performance-optimizer` | Speed | Lighthouse, Profiling |
-| `seo-specialist` | SEO | Meta, Schema, Rankings |
-| `documentation-writer` | Docs | README, API docs |
-| `debugger` | Debug | Error analysis |
+```markdown
+## Meta-Agent Verification
+
+- [ ] **assessor** was invoked for risk assessment
+- [ ] **recovery** saved checkpoint before risky operations
+- [ ] **critic** resolved any agent conflicts
+- [ ] **learner** will log outcome (pending completion)
+- [ ] **observability** are being collected (AutopilotMetrics)
+```
+
+> **VIOLATION:** Skipping meta-agent hooks = Autopilot incomplete.
+> All hooks are logged and auditable.
 
 ---
 
+
+
+
 ## Execution Phases
 
 ### Phase 1: Architecture (Sequential)
 
-| Step | Agent | Action |
-|------|-------|--------|
-| 1 | `project-planner` | Create PLAN.md |
-| 2 | `explorer-agent` | Codebase discovery (optional) |
+| Step | Agent             | Action                        |
+| ---- | ----------------- | ----------------------------- |
+| 1    | `project-planner` | Create PLAN.md                |
+| 2    | `explorer-agent`  | Codebase discovery (optional) |
+
+> 🔴 **NO OTHER AGENTS during planning!**
+
+**â›” CHECKPOINT: User approval required before Phase 2**
+
+---
+
+## 🎨 Design System (Required for UI Apps)
+
+> **CRITICAL:** Generate design system BEFORE building UI components.
+
+### Step 1: Generate Design System
+
+// turbo
+
+```bash
+node .agent/skills/studio/scripts-js/search.js "<app_type> <style> <keywords>" --design-system -p "<Project Name>"
+```
+
+**Examples:**
+
+```bash
+# Weather app
+node .agent/skills/studio/scripts-js/search.js "weather dashboard glassmorphism modern" --design-system -p "Weather App"
+
+# E-commerce
+node .agent/skills/studio/scripts-js/search.js "ecommerce beauty elegant soft" --design-system -p "Beauty Store"
+
+# SaaS
+node .agent/skills/studio/scripts-js/search.js "saas fintech professional dark" --design-system -p "PayFlow"
+```
+
+### Step 2: Apply Design Tokens
 
-> 🔴 **NO OTHER AGENTS during planning!**
+Use the generated design system to define:
 
-**⛔ CHECKPOINT: User approval required before Phase 2**
+| Token          | Source                          | Usage                          |
+| -------------- | ------------------------------- | ------------------------------ |
+| **Colors**     | `--design-system` output        | `bg-primary`, `text-accent`    |
+| **Typography** | Font pairing recommendation     | `font-heading`, `font-body`    |
+| **Effects**    | Visual effects (glass, shadows) | `backdrop-blur`, `shadow-glow` |
+
+### Anti-Patterns to Avoid
+
+| ❌ Don't                   | ✅ Do                             |
+| -------------------------- | --------------------------------- |
+| Use emoji as icons (🎨 🚀) | Use SVG icons (Lucide, Heroicons) |
+| Generic purple gradients   | Use curated color palette         |
+| Random font choices        | Use recommended font pairing      |
+| Scale transforms on hover  | Use opacity/color transitions     |
+
+> **Reference:** See `/studio` workflow for detailed guidance.
+
+---
 
 ### Phase 2: Parallel Execution
 
@@ -120,27 +358,49 @@ graph TD
     D2 --> E
 ```
 
-| Parallel Group | Agents |
-|----------------|--------|
-| Foundation | database-architect, security-auditor |
-| Core | backend-specialist, frontend-specialist |
-| Polish | test-engineer, devops-engineer |
+| Parallel Group | Agents                                  |
+| -------------- | --------------------------------------- |
+| Foundation     | database-architect, security-auditor    |
+| Core           | backend-specialist, frontend-specialist |
+| Polish         | test-engineer, devops-engineer          |
 
 ### Phase 3: Verification
+
 // turbo
+
 ```bash
-python .agent/skills/vulnerability-scanner/scripts/security_scan.py .
-python .agent/skills/lint-and-validate/scripts/lint_runner.py .
+python .agent/skills/security-scanner/scripts/security_scan.py .
+python .agent/skills/code-quality/scripts/lint_runner.py .
 ```
 
 ---
 
+## 🚀 Auto-Execution Policy
+
+**This workflow runs in CONTINUOUS EXECUTION MODE.**
+
+Once user approves plan:
+
+1. **All commands auto-run** (`SafeToAutoRun: true`)
+2. **No mid-phase prompts** (see CONTINUOUS_EXECUTION_POLICY.md)
+3. **Only stop for:**
+   - Blocking errors
+   - Decision forks
+   - Plan completion
+
+**Policy:** User approval = Full trust for execution
+
+**Reference:** `.agent/CONTINUOUS_EXECUTION_POLICY.md`
+
+---
+
 ## Context Passing (MANDATORY)
 
 When invoking ANY sub-agent, include:
 
 ```markdown
 **CONTEXT:**
+
 - Original Request: [Full user request]
 - Decisions Made: [All user answers]
 - Previous Agent Work: [Summary of completed work]
@@ -149,35 +409,106 @@ When invoking ANY sub-agent, include:
 **TASK:** [Specific task for this agent]
 ```
 
-> ⚠️ **VIOLATION:** Invoking agent without context = wrong assumptions!
+> ⚠️ **VIOLATION:** Invoking agent without context = wrong assumptions!
+
+---
+
+## Command Execution Rules
+
+When invoking commands during autopilot:
+
+```javascript
+run_command({
+  CommandLine: "npm install ...",
+  SafeToAutoRun: true, // ✅ REQUIRED in autopilot mode
+  WaitMsBeforeAsync: 10000,
+  Cwd: "...",
+});
+```
+
+**Rules:**
+
+- ✅ **Always set `SafeToAutoRun: true`** after plan approval
+- ❌ **Never use `SafeToAutoRun: false`** in autopilot execution
+- 📋 **Reference:** See Auto-Execution Policy above
+
+---
+
+## 📦 Scaffold Scripts (Minimize Accept Prompts)
+
+Pre-approved scripts bundle multiple commands to reduce Accept prompts.
+
+### Available Scripts
+
+| Script               | Purpose                                   | Saves         |
+| -------------------- | ----------------------------------------- | ------------- |
+| `scaffold-nextjs.js` | Create Next.js project + deps + structure | 3+ → 1 prompt |
+
+### Usage
+
+```bash
+# Instead of multiple npx/npm commands (3+ Accept prompts):
+npx create-next-app...  # ← Accept
+npm install zustand...  # ← Accept
+npm run dev...          # ← Accept
+
+# Use scaffold script (1 Accept prompt):
+node .agent/scripts-js/scaffold-nextjs.js my-app --deps zustand,framer-motion --start
+```
+
+### scaffold-nextjs.js Options
+
+```bash
+node .agent/scripts-js/scaffold-nextjs.js <name> [options]
+
+Options:
+  --port <number>   Dev server port (default: 3000)
+  --deps <packages> Extra npm packages (comma-separated)
+  --no-install      Skip installing extra dependencies
+  --start           Auto-start dev server after creation
+```
+
+### Example: Weather App
+
+```bash
+node .agent/scripts-js/scaffold-nextjs.js weather-app --port 3001 --start
+```
+
+**Result:**
+
+- ✅ Next.js 15 + TypeScript + Tailwind
+- ✅ Auto-created: components/, lib/, hooks/, types/
+- ✅ Dev server running on port 3001
+- ✅ Only 1 Accept prompt!
 
 ---
 
 ## Output Format
 
 ```markdown
-## 🎼 Autopilot Report
+## 🎼 Autopilot Report
 
 ### Mission
+
 [Original task summary]
 
 ### Agent Coordination
 
-| Agent | Task | Duration | Status |
-|-------|------|----------|--------|
-| `project-planner` | Task breakdown | 2m | ✅ Complete |
-| `database-architect` | Schema design | 3m | ✅ Complete |
-| `backend-specialist` | API routes | 5m | ✅ Complete |
-| `frontend-specialist` | UI components | 7m | ✅ Complete |
-| `test-engineer` | E2E tests | 4m | ✅ Complete |
+| Agent                 | Task           | Duration | Status      |
+| --------------------- | -------------- | -------- | ----------- |
+| `project-planner`     | Task breakdown | 2m       | ✅ Complete |
+| `database-architect`  | Schema design  | 3m       | ✅ Complete |
+| `backend-specialist`  | API routes     | 5m       | ✅ Complete |
+| `frontend-specialist` | UI components  | 7m       | ✅ Complete |
+| `test-engineer`       | E2E tests      | 4m       | ✅ Complete |
 
 ### Verification Results
 
-| Script | Result |
-|--------|--------|
-| security_scan.py | ✅ No vulnerabilities |
-| lint_runner.py | ✅ No errors |
-| Tests | ✅ 28/28 passed |
+| Script           | Result                |
+| ---------------- | --------------------- |
+| security_scan.py | ✅ No vulnerabilities |
+| lint_runner.py   | ✅ No errors          |
+| Tests            | ✅ 28/28 passed       |
 
 ### Deliverables
 
@@ -189,15 +520,18 @@ When invoking ANY sub-agent, include:
 - [x] Preview running
 
 ### Summary
+
 Built complete [app type] with [X] files across [Y] agents.
 Total execution time: [Z] minutes.
 
 ---
 
 ### Preview
-🌐 http://localhost:3000
+
+🌐 http://localhost:3000
 
 ### Next Steps
+
 - [ ] Review the code
 - [ ] Test user flows
 - [ ] `/launch` when ready
@@ -217,20 +551,312 @@ Total execution time: [Z] minutes.
 
 ---
 
-## Exit Gate
+## â›” MANDATORY: Problem Verification Before Completion
+
+> **CRITICAL:** This check MUST be performed before ANY `notify_user` call or task completion.
+
+### Step 1: Check IDE Problems
+
+```
+Before marking task complete:
+1. Read @[current_problems] from IDE
+2. Count errors and warnings
+3. If count > 0 → DO NOT COMPLETE → Go to Step 2
+4. If count = 0 → Proceed to Exit Gate
+```
+
+### Step 2: Auto-Fix Protocol
+
+```javascript
+// Pseudo-code for problem handling
+const problems = getIDEProblems();
+
+if (problems.length > 0) {
+  for (const problem of problems) {
+    if (isAutoFixable(problem)) {
+      autoFix(problem); // Fix: imports, types, lint errors
+    } else {
+      escalateToUser(problem); // Can't auto-fix: notify user
+      return; // DO NOT mark complete
+    }
+  }
+
+  // Re-check after auto-fix
+  const remainingProblems = getIDEProblems();
+  if (remainingProblems.length > 0) {
+    escalateToUser(remainingProblems);
+    return; // DO NOT mark complete
+  }
+}
+
+// Only now proceed to completion
+proceedToExitGate();
+```
+
+### Auto-Fixable Issues
+
+| Type                | Example                          | Fix Method                |
+| ------------------- | -------------------------------- | ------------------------- |
+| **Missing import**  | `ReactNode` not imported         | Add import statement      |
+| **JSX namespace**   | `Cannot find namespace 'JSX'`    | Import from 'react'       |
+| **Unused variable** | `'x' is declared but never used` | Remove or prefix with `_` |
+| **Lint errors**     | Semicolons, spacing              | Run prettier/eslint --fix |
+| **Type errors**     | Simple type mismatches           | Add type assertion        |
+
+### Non-Fixable Issues (Escalate)
+
+| Type                 | Example               | Action                        |
+| -------------------- | --------------------- | ----------------------------- |
+| **Logic errors**     | Wrong business logic  | Notify user, ask for guidance |
+| **Breaking changes** | API contract broken   | Notify user, block completion |
+| **Missing deps**     | Package not installed | Ask user to install           |
+
+> **Rule:** Never call `notify_user` with task completion if `@[current_problems]` shows errors.
+
+---
+
+## Exit Gate (Enhanced)
+
+Before completing, verify ALL criteria:
+
+### Mandatory Checks
+
+| Check               | Target | How to Verify               |
+| ------------------- | ------ | --------------------------- |
+| **Agent Count**     | ≥3     | Count unique agents invoked |
+| **IDE Problems**    | 0      | Check `@[current_problems]` |
+| **Security Scan**   | Pass   | `security_scan.py` output   |
+| **Preview Running** | Yes    | `npm run dev` active        |
+
+### Quality observability
+
+| Metric             | Target           | How to Measure     |
+| ------------------ | ---------------- | ------------------ |
+| **Files Created**  | Per plan         | Count deliverables |
+| **Lint Errors**    | 0                | ESLint output      |
+| **Type Errors**    | 0                | TypeScript check   |
+| **Execution Time** | <5min (standard) | Compare start/end  |
+
+### Exit Decision Tree
+
+```
+IDE Problems > 0?
+├── YES → Auto-fix if possible, else notify user
+└── NO → Continue
+
+Security Scan failed?
+├── YES → Block completion, notify user
+└── NO → Continue
+
+All planned deliverables complete?
+├── YES → Generate report, notify user
+└── NO → Continue execution
+```
+
+> **Rule:** If ANY check fails → DO NOT mark complete. Fix or escalate.
+
+---
+
+## 📊 Performance Benchmarks
+
+### Real-World Baseline (TodoList Execution)
+
+| Phase     | Operation          | Actual Time | Target    | Status        |
+| --------- | ------------------ | ----------- | --------- | ------------- |
+| Setup     | Project creation   | 23s         | <60s      | ✅ Pass       |
+| Setup     | Dependency install | 11s         | <45s      | ✅ Pass       |
+| Build     | Types/Store        | ~5s         | <10s      | ✅ Pass       |
+| Build     | Components (5)     | ~30s        | <60s      | ✅ Pass       |
+| Build     | Pages/Layout       | ~10s        | <20s      | ✅ Pass       |
+| Verify    | Problem check      | <2s         | <5s       | ✅ Pass       |
+| Verify    | Auto-fix           | <5s         | <10s      | ✅ Pass       |
+| **Total** | **End-to-end**     | **~8min**   | **<5min** | ⚠️ Exceeded\* |
+
+_\*Exceeded due to user approval prompts. After turbo-all fix, expect <5min._
+
+### Benchmark by App Type
+
+| App Type       | Agents | Expected Time | Files | Example               |
+| -------------- | ------ | ------------- | ----- | --------------------- |
+| **Simple**     | 2-3    | <3 min        | <10   | Landing page          |
+| **Standard**   | 3-4    | 3-5 min       | 10-20 | TodoList, Blog        |
+| **Complex**    | 5+     | 5-10 min      | 20-50 | E-commerce, Dashboard |
+| **Enterprise** | 7+     | 10-20 min     | 50+   | Full SaaS             |
+
+### Time Breakdown by Phase
+
+```
+Typical autopilot execution:
+┌─────────────────────────────────────────────┐
+│ Phase 1: Planning (15-20%)                  │
+│ ████████░░░░░░░░░░░░░░░░░░░░░░░░░░ (~1min) │
+├─────────────────────────────────────────────┤
+│ Phase 2: Execution (60-70%)                 │
+│ ████████████████████████████░░░░░░ (~3min) │
+├─────────────────────────────────────────────┤
+│ Phase 3: Verification (15-20%)              │
+│ ████████░░░░░░░░░░░░░░░░░░░░░░░░░░ (~1min) │
+└─────────────────────────────────────────────┘
+```
+
+---
+
+## 💰 Cost Estimation
+
+### Formula
+
+```javascript
+function estimateAutopilotCost(complexity, agents) {
+  const baseTime = {
+    simple: 3, // minutes
+    standard: 5,
+    complex: 10,
+    enterprise: 20,
+  };
+
+  const agentMultiplier = Math.max(1, agents / 3); // Normalized to 3 agents
+  const base = baseTime[complexity];
+
+  return {
+    optimistic: Math.round(base * 0.8), // Best case
+    likely: Math.round(base * agentMultiplier), // Expected
+    pessimistic: Math.round(base * 1.5 * agentMultiplier), // Worst case
+  };
+}
+
+// Example: Standard app with 4 agents
+estimateAutopilotCost("standard", 4);
+// → { optimistic: 4, likely: 7, pessimistic: 10 } minutes
+```
+
+### Quick Reference Table
+
+| Complexity | Agents | Best Case | Expected | Worst Case |
+| ---------- | ------ | --------- | -------- | ---------- |
+| Simple     | 3      | 2 min     | 3 min    | 5 min      |
+| Standard   | 4      | 4 min     | 7 min    | 10 min     |
+| Complex    | 5      | 8 min     | 12 min   | 18 min     |
+| Enterprise | 7      | 16 min    | 25 min   | 35 min     |
+
+### What's NOT Included
+
+- User thinking/approval time
+- Network latency variations
+- Unexpected complexity discoveries
+- External API calls
+
+---
+
+## 🎛️ Monitoring Dashboard
+
+### Real-Time View (Conceptual)
+
+```
+┌──────────────────────────────────────────────────────┐
+│ 🎼 AUTOPILOT: FAANG TodoList                         │
+│ Status: 🔄 Phase 2/3 - Building                      │
+│ Elapsed: 2m 34s / ~5m estimated                      │
+│ Progress: ████████████████░░░░░░░░░░ 65%             │
+└──────────────────────────────────────────────────────┘
+
+┌─────────────────────────┬────────────────────────────┐
+│ 📋 PHASE PROGRESS       │ 👥 AGENT STATUS            │
+├─────────────────────────┼────────────────────────────┤
+│ ✅ Phase 1: Planning    │ ✅ project-planner (1m 12s)│
+│    └─ 1m 12s            │ 🔄 frontend-specialist     │
+│ 🔄 Phase 2: Execution   │    └─ active (45s)         │
+│    └─ 1m 22s / ~3m      │ ⏳ backend-specialist      │
+│ ⏳ Phase 3: Verification │ ⏳ test-engineer           │
+│    └─ Not started        │                            │
+└─────────────────────────┴────────────────────────────┘
+
+┌──────────────────────────────────────────────────────┐
+│ 📊 QUALITY observability                                   │
+├──────────────────────────────────────────────────────┤
+│ IDE Problems:  2 → 0 (auto-fixed)       ✅           │
+│ Security Scan: Pending                  ⏳           │
+│ Lint Check:    Pending                  ⏳           │
+│ Type Check:    0 errors                 ✅           │
+│ Success Rate:  95% (historical)         📈           │
+└──────────────────────────────────────────────────────┘
+
+┌──────────────────────────────────────────────────────┐
+│ 💾 RESOURCE USAGE                                    │
+├──────────────────────────────────────────────────────┤
+│ Memory: ████████░░░░░░░░ 234 MB / 500 MB             │
+│ Disk:   ████░░░░░░░░░░░░░ 1.2 GB / 2 GB              │
+│ Files:  11 created, 0 modified, 0 deleted            │
+│ Agents: 2/4 active, 2 queued                         │
+└──────────────────────────────────────────────────────┘
+```
+
+### observability to Track
+
+| Metric         | Update Frequency | Visibility    |
+| -------------- | ---------------- | ------------- |
+| Phase progress | Per phase        | Always        |
+| Agent status   | Per action       | Always        |
+| IDE problems   | Per file write   | On change     |
+| Time elapsed   | Per second       | Always        |
+| Resource usage | Per 10s          | Expanded view |
+
+---
+
+## 🔧 Troubleshooting Guide
+
+### Common Issues & Solutions
+
+| Issue                   | Symptoms                      | Solution                        |
+| ----------------------- | ----------------------------- | ------------------------------- |
+| **Approval prompts**    | "Accept?" after plan approval | Add `// turbo-all` annotation   |
+| **IDE problems at end** | Warnings/errors on completion | Check auto-fix coverage         |
+| **Slow execution**      | >10min for standard app       | Check agent parallelization     |
+| **Security scan fail**  | Vulnerabilities detected      | Review dependencies, update     |
+| **npm install fail**    | Package conflicts             | Clear node_modules, retry       |
+| **TypeScript errors**   | Type check fails              | Fix or use `as unknown as Type` |
+
+### Debugging Checklist
+
+When autopilot fails or produces unexpected results:
+
+```markdown
+## Debug Checklist
+
+- [ ] Check `@[current_problems]` for IDE issues
+- [ ] Review console output for errors
+- [ ] Verify target directory is writable
+- [ ] Check npm/node versions are compatible
+- [ ] Review PLAN.md for missed requirements
+- [ ] Check if all agents were invoked
+- [ ] Verify preview server is running
+- [ ] Review final report for observability
+```
+
+### Recovery Procedures
+
+**Partial Failure (some phases complete):**
 
-Before completing, verify:
+1. Identify last successful phase
+2. Review what failed and why
+3. Fix issue manually or retry phase
+4. Continue from where it stopped
 
-- [ ] **Agent Count:** invoked_agents >= 3
-- [ ] **Scripts Executed:** At least security_scan.py ran
-- [ ] **Report Generated:** All agents listed with status
-- [ ] **Deliverables:** All planned items completed
+**Complete Failure (nothing works):**
 
-> If any check fails → DO NOT mark complete. Invoke more agents or run scripts.
+1. Save any generated files
+2. Delete target directory
+3. Review plan for issues
+4. Restart autopilot with fixes
+
+**Rollback (need to undo):**
+
+1. If git initialized: `git checkout -- .`
+2. If no git: Delete target directory
+3. No partial rollback - all or nothing
 
 ---
 
-## 🔗 Workflow Chain
+## 🔗 Workflow Chain
 
 ```mermaid
 graph LR
@@ -241,14 +867,51 @@ graph LR
     style A fill:#3b82f6
 ```
 
-| /autopilot integrates | Purpose |
-|-----------------------|---------|
-| `/architect` | Planning phase |
-| `/build` | Building phase |
-| `/validate` | Testing phase |
-| `/launch` | Deployment phase |
+| /autopilot integrates | Purpose          |
+| --------------------- | ---------------- |
+| `/architect`          | Planning phase   |
+| `/build`              | Building phase   |
+| `/validate`           | Testing phase    |
+| `/launch`             | Deployment phase |
+
+---
+
+## 📊 Final observability Report Template
+
+Include this in every autopilot completion:
+
+```markdown
+## 📊 Execution observability
+
+| Metric            | Target | Actual | Status |
+| ----------------- | ------ | ------ | ------ |
+| Agents invoked    | ≥3     | [X]    | ✅/❌  |
+| IDE problems      | 0      | [X]    | ✅/❌  |
+| Phase transitions | <2s    | [X]s   | ✅/❌  |
+| Total execution   | <5min  | [X]m   | ✅/❌  |
+| Files created     | [plan] | [X]    | ✅/❌  |
+| Auto-fix rate     | >85%   | [X]%   | ✅/❌  |
+
+### SLO Compliance
+
+- ✅ First-time success: YES/NO
+- ✅ Zero manual fixes needed: YES/NO
+- ✅ All deliverables complete: YES/NO
+```
+
+---
 
 **Handoff:**
+
 ```markdown
-Autopilot complete. All agents finished. Preview running.
+Autopilot complete. All agents finished. Preview running at localhost:3000.
+
+See observability report above for quality verification.
 ```
+
+---
+
+**Version:** 2.0.0  
+**Policy Reference:** `.agent/CONTINUOUS_EXECUTION_POLICY.md`  
+**Applies to:** All autopilot executions
+