proagents 1.6.12 → 1.6.15

package/.proagents/AI_INSTRUCTIONS.md

@@ -2,6 +2,49 @@
 
 This project uses ProAgents - an AI-agnostic development workflow framework.
 
+---
+
+## ⚠️ MANDATORY: Auto-Sync on EVERY pa: Command
+
+**BEFORE executing ANY `pa:` command, AI MUST run these steps FIRST:**
+
+```bash
+# STEP 1: Load context (REQUIRED - do this FIRST)
+cat .proagents/worklog/_context.md 2>/dev/null
+cat .proagents/changelog/_recent.md 2>/dev/null | head -30
+tail -10 .proagents/activity.log 2>/dev/null
+```
+
+**This applies to ALL commands including:**
+- `pa:fix "bug"` - Load context FIRST, then fix
+- `pa:feature "name"` - Load context FIRST, then start feature
+- `pa:test` - Load context FIRST, then run tests
+- `pa:doc` - Load context FIRST, then generate docs
+- **EVERY pa: command** - Load context FIRST
+
+**AFTER executing ANY `pa:` command that changes files, AI MUST:**
+
+```bash
+# Update worklogs and changelogs
+# 1. Prepend to .proagents/changelog/_recent.md
+# 2. Update .proagents/worklog/_context.md
+# 3. Update .proagents/changelog/modules/[module].md
+# 4. Log to .proagents/activity.log
+```
+
+**Example flow for `pa:fix "login bug"`:**
+```
+1. AI reads _context.md, _recent.md, activity.log  ← AUTO-SYNC FIRST
+2. AI analyzes and fixes the bug                    ← DO THE WORK
+3. AI updates _recent.md with fix details           ← LOG THE CHANGE
+4. AI updates _context.md with current state        ← UPDATE CONTEXT
+5. AI logs to activity.log                          ← LOG ACTIVITY
+```
+
+**Note:** When updating `_recent.md` or `_context.md`, first read the corresponding `*.template.md` file to use the latest format. See "Using Template Reference Files" section below.
+
+---
+
 ## CRITICAL PRINCIPLE: ALWAYS DO, NEVER JUST TELL
 
 **ProAgents automates development. The AI must EXECUTE actions, not give instructions for users to follow.**
@@ -31,24 +74,24 @@ This project uses ProAgents - an AI-agnostic development workflow framework.
 
 **Multiple AI tools may work on this project simultaneously. They do NOT share context.**
 
-### AUTOMATIC Context Loading (No Manual pa:sync Needed)
+### AUTOMATIC Context Loading (EVERY pa: Command)
 
-**On FIRST pa: command of session, AI AUTOMATICALLY does:**
+**On EVERY `pa:` command, AI MUST FIRST load context:**
 
 ```bash
-# 1. Load context (replaces manual pa:sync)
+# MANDATORY FIRST STEP - Run before ANY pa: command
 cat .proagents/worklog/_context.md 2>/dev/null
 cat .proagents/changelog/_recent.md 2>/dev/null | head -30
 tail -10 .proagents/activity.log 2>/dev/null
 
-# 2. Check for conflicts
+# Also check for conflicts
 cat .proagents/.active-files 2>/dev/null
-
-# 3. Validate previous logging
-git diff --name-only HEAD~3 2>/dev/null | wc -l
 ```
 
-**User does NOT need to type pa:sync - it happens automatically.**
+**This is NOT optional. This is NOT just for "first command".**
+**EVERY pa: command = Load context FIRST, then execute.**
+
+**User does NOT need to type pa:sync - AI does it automatically on EVERY pa: command.**
 
 ### Before ANY `pa:` command:
 
@@ -114,6 +157,37 @@ AI immediately updates these files (no user prompt needed):
 4. .proagents/changelog/modules/X.md   ← Based on file path
 ```
 
+### Using Template Reference Files
+
+**IMPORTANT:** Template files (`*.template.md`, `*.template.json`) provide the latest format for user data files.
+
+| User Data File (EDIT THIS) | Template Reference (READ ONLY) |
+|----------------------------|-------------------------------|
+| `changelog/_recent.md` | `changelog/_recent.template.md` |
+| `worklog/_context.md` | `worklog/_context.template.md` |
+| `worklog/ai-stats.json` | `worklog/ai-stats.template.json` |
+| `active-features/_index.json` | `active-features/_index.template.json` |
+| `.learning/global/user-preferences.json` | `.learning/global/user-preferences.template.json` |
+| `.learning/global/common-patterns.json` | `.learning/global/common-patterns.template.json` |
+
+**How to use:**
+1. **Read the template FIRST** before creating/updating user data files
+2. **Apply the template format** to user data (structure, fields, sections)
+3. **NEVER modify template files** - they are READ ONLY (updated by framework)
+
+**Why templates exist:**
+- Templates get updated when ProAgents upgrades (new fields, better structure)
+- User data files are preserved during upgrades (your content stays safe)
+- AI always has access to the latest format by reading templates
+
+**Common mistake to avoid:**
+```
+WRONG: Edit _recent.template.md directly
+RIGHT: Read _recent.template.md, then edit _recent.md using that format
+```
+
+---
+
 ### Auto-Log Format for _recent.md:
 
 After editing `src/auth/login.ts`, AI prepends to `_recent.md`:
@@ -189,6 +263,81 @@ When removing or correcting previous changes, also log:
 
 ---
 
+## Smart Context Features (AUTO)
+
+### 1. Quick Summary (Top of _context.md)
+
+After EVERY change, AI regenerates Quick Summary (max 5 lines):
+
+```
+## Quick Summary
+Last: Claude fixed login bug #123 (2 hrs ago)
+Active: user-auth feature (70%)
+Pending: 2 tasks
+Tests: ✓ Passing
+```
+
+### 2. Context Changed Alert
+
+On first pa: command, check if another AI worked:
+
+```bash
+# AI compares last session vs _recent.md
+LAST_AI=$(grep "Last AI:" .proagents/worklog/_context.md | cut -d: -f2)
+RECENT_AI=$(head -20 .proagents/changelog/_recent.md | grep "AI:" | head -1)
+```
+
+If different AI made changes, show:
+```
+⚠️ CONTEXT CHANGED
+Gemini made 3 changes 1 hour ago:
+- Fixed API endpoint
+- Updated auth flow
+Review before continuing? [Y/n]
+```
+
+### 3. Test Status Auto-Update
+
+After ANY test command, update _context.md:
+
+```
+## Test Status
+Status: ✓ 45 passing | ✗ 2 failing
+Last Run: 2024-03-11 15:30
+Failing: auth.test.ts (line 45)
+```
+
+### 4. Feature Progress Auto-Calculate
+
+```
+## Feature Progress
+Feature: user-auth
+Progress: ████████░░ 80%
+Completed: 8/10 tasks
+```
+
+Calculate: (completed tasks / total tasks) × 100
+
+### 5. Auto-Archive Old Logs (7+ days)
+
+```bash
+mkdir -p .proagents/worklog/archive
+# Move files older than 7 days
+find .proagents/worklog -maxdepth 1 -name "2*.md" -mtime +7 -exec mv {} .proagents/worklog/archive/ \;
+```
+
+### 6. AI Performance Stats
+
+Update `worklog/ai-stats.json` after each session:
+
+```json
+{
+  "Claude": { "sessions": 15, "tasks": 45, "reverts": 2 }
+}
+```
+
+---
+
 ## Cross-AI Session Tracking (AUTOMATIC)
 
 **All session tracking is AUTOMATIC. User does NOT need to run these commands manually.**
@@ -201,6 +350,10 @@ When removing or correcting previous changes, also log:
 | Log changes | After each edit | **None** - Auto |
 | Update _context.md | After each edit | **None** - Auto |
 | Check conflicts | Before editing files | **None** - Auto |
+| Update Quick Summary | After each edit | **None** - Auto |
+| Context changed alert | First pa: command | **None** - Auto |
+| Update test status | After tests | **None** - Auto |
+| Update AI stats | After session | **None** - Auto |
 
 ### On First pa: Command (Auto-Sync):
 
@@ -402,11 +555,14 @@ Auto-detect issue numbers from user input and include in changelog:
 ### Collaboration
 | Command | Action |
 |---------|--------|
-| `pa:activity` | Show AI activity log |
+| `pa:activity` | **Show ALL AI platforms activity** (not just yours!) |
 | `pa:lock` | Show lock status |
 | `pa:handoff` | Create handoff notes |
 | `pa:feedback "text"` | Log feedback for AI learning |
 | `pa:decision "title"` | Log architectural decision |
+| `pa:learn "pattern"` | Teach AI a pattern → `.learning/patterns.json` |
+| `pa:forget "pattern"` | Remove learned pattern |
+| `pa:learning` | Show learning report (patterns, corrections, stats) |
 
 ### Cross-AI Continuity (CRITICAL)
 | Command | Action |
@@ -436,7 +592,7 @@ Auto-detect issue numbers from user input and include in changelog:
 |---------|--------|
 | `pa:history` | Show command history (read activity.log) |
 | `pa:progress` | Show feature progress |
-| `pa:activity` | Show recent AI activity |
+| `pa:activity` | **Show ALL AI platforms activity** (cross-AI visibility) |
 
 ---
 
@@ -471,34 +627,189 @@ When user types a `pa:` command:
 
 ---
 
+## pa:help Execution
+
+**Show comprehensive help with examples and categories.**
+
+**Command variations:**
+```
+pa:help                        # Full help with examples
+pa:help <command>              # Detailed help for specific command
+pa:help --quick                # Quick reference only
+pa:help --examples             # Show more examples
+```
+
+**Output format:**
+```
+ProAgents Help
+══════════════════════════════════════════════════════════
+
+🚀 Quick Start Examples
+───────────────────────
+  pa:feature "user authentication"    Start a new feature
+  pa:fix "login button not working"   Quick bug fix
+  pa:status                           Check current progress
+  pa:resume                           Continue where you left off
+
+📋 Common Workflows
+───────────────────
+  Feature Development:
+    pa:feature → pa:analyze → pa:design → pa:implement → pa:test → pa:deploy
+
+  Bug Fix (fast track):
+    pa:fix "description" → auto-analyze → fix → test → done
+
+  Returning to Work:
+    pa:sync → pa:resume → continue working
+
+📂 Commands by Category
+───────────────────────
+
+  Core Commands:
+    pa:feature "name"     Start new feature with full workflow
+    pa:fix "bug"          Quick bug fix mode
+    pa:status             Show current progress (enhanced)
+    pa:help               Show this help
+
+  Workflow Phases:
+    pa:analyze            Deep codebase analysis
+    pa:requirements       Gather requirements
+    pa:design             UI/Architecture design
+    pa:plan               Create implementation plan
+    pa:implement          Execute implementation
+    pa:test               Create and run tests
+    pa:review             Code review workflow
+    pa:doc                Generate documentation
+    pa:deploy             Deployment preparation
+
+  Progress & History:
+    pa:progress           Visual progress (all features)
+    pa:status             Detailed status (current feature)
+    pa:activity           All AI activity with filters
+    pa:history            Command history with stats
+
+  Cross-AI Collaboration:
+    pa:sync               Load context + visual diff
+    pa:resume             Quick resume with next action
+    pa:handoff            Create handoff notes
+    pa:conflict-check     Check for conflicts
+
+  Learning & Tracking:
+    pa:feedback "text"    Log feedback for AI learning
+    pa:decision "title"   Log architectural decision
+    pa:error "desc"       Log error and solution
+    pa:learn "pattern"    Teach AI a pattern
+    pa:forget "pattern"   Remove learned pattern
+    pa:learning           Show learning report
+
+💡 Tips
+───────
+  • Use aliases: pa:f (feature), pa:s (status), pa:h (help)
+  • AI auto-syncs on every command - no manual sync needed
+  • Changes are auto-logged to changelog
+
+Type "pa:help <command>" for detailed help on a specific command.
+Example: pa:help feature
+```
+
+**For specific command help (pa:help feature):**
+```
+pa:feature - Start New Feature
+══════════════════════════════════════════════════════════
+
+Usage:
+  pa:feature "feature name"
+  pa:feature "user authentication"
+
+What it does:
+  1. Creates feature folder in .proagents/active-features/
+  2. Analyzes codebase for relevant files
+  3. Guides through full workflow phases
+  4. Tracks progress and tasks
+
+Workflow phases:
+  Analysis → Requirements → Design → Planning →
+  Implementation → Testing → Review → Documentation → Deployment
+
+Options:
+  pa:feature "name" --fast      Skip optional phases
+  pa:feature "name" --no-tests  Skip test generation
+
+Related commands:
+  pa:feature-list       List all features
+  pa:feature-complete   Mark feature as complete
+  pa:status             Check feature status
+```
+
+---
+
 ## pa:history Execution
 
 **AI MUST read and display the actual log file content.**
 
+**Command variations:**
+```
+pa:history                     # All recent history (default)
+pa:history --today             # Today's commands only
+pa:history --ai Claude         # Filter by AI platform
+pa:history --command pa:fix    # Filter by command type
+pa:history --search "auth"     # Search in descriptions
+pa:history --stats             # Show command statistics
+```
+
+**Steps:**
 1. Read `./.proagents/activity.log`
-2. Filter out header lines (lines starting with #)
-3. Display entries (most recent first)
+2. Filter based on options (if provided)
+3. Display entries with grouping and icons
 4. NEVER say "no commands yet" without reading the file
 
 ```bash
 # AI runs this:
-grep -v "^#" .proagents/activity.log | grep -v "^$" | grep -v "Activity Log Start" | tail -30
+grep -v "^#" .proagents/activity.log | grep -v "^$" | grep -v "Activity Log Start" | tail -50
 ```
 
-If no entries found after filtering, show:
+**Enhanced output format:**
 ```
-No pa: commands logged yet.
-Run some commands, then try pa:history again.
+Command History
+══════════════════════════════════════════════════════════
+
+Today (12 commands)
+───────────────────
+[16:00] [Gemini]  ✅ pa:test - 12 tests passed
+[15:30] [Cursor]  🐛 pa:fix - Fixed login validation
+[15:15] [Claude]  🔍 pa:analyze - Analyzed 45 files
+[15:10] [Claude]  ✨ pa:feature - Started "user-auth"
+
+Yesterday (8 commands)
+──────────────────────
+[18:00] [ChatGPT] 📝 pa:doc - Updated README
+[16:30] [Claude]  🔨 pa:implement - Created UserService
+
+──────────────────────────────────────────────────────────
+Stats: 20 commands | 4 AIs | Most used: pa:implement (6)
 ```
 
-**Output format:**
+**With --stats option:**
 ```
-Command History
-═══════════════
-[2024-03-06 16:00] [Gemini] pa:logs - Captured 50 entries
-[2024-03-06 15:30] [Cursor] pa:test - 12 tests passed
-[2024-03-06 15:15] [Claude] pa:analyze - Analyzed 45 files
-[2024-03-06 15:10] [Claude] pa:feature - Started "user-auth"
+Command Statistics
+══════════════════════════════════════════════════════════
+
+By Command:
+  pa:implement   ████████████████  12
+  pa:fix         ████████          6
+  pa:test        ██████            4
+  pa:analyze     ████              3
+
+By AI Platform:
+  Claude         ████████████████  14
+  Cursor         ████████          7
+  Gemini         ████              3
+  ChatGPT        ██                1
+
+By Day:
+  Today          ████████████████  12
+  Yesterday      ████████          8
+  Earlier        ████              5
 ```
 
 ---
@@ -509,25 +820,50 @@ Command History
 
 1. Read `./.proagents/active-features/_index.json`
 2. Read each feature's `status.json`
-3. Calculate and display progress
+3. Read `./.proagents/activity.log` for last AI info
+4. Calculate progress and display enhanced output
 
 ```bash
 # AI runs this:
 cat .proagents/active-features/_index.json
 cat .proagents/active-features/feature-*/status.json
+cat .proagents/activity.log | tail -50
 ```
 
-**Output format:**
+**Output format (enhanced):**
+```
+Feature Progress
+════════════════════════════════════════════════════════════
+
+user-auth       ████████████████░░░░  80%  (8/10 tasks)  3d  Claude
+                └─ Analysis ✓ → Requirements ✓ → Design ✓ → Implementation ✓ → Testing ● → Review ○
+
+dashboard       ████████░░░░░░░░░░░░  40%  (4/10 tasks)  5d  Cursor
+                └─ Analysis ✓ → Requirements ✓ → Design ● → Implementation ○ → Testing ○ → Review ○
+
+notifications   ██░░░░░░░░░░░░░░░░░░  10%  (1/10 tasks)  1d  ChatGPT
+                └─ Analysis ● → Requirements ○ → Design ○ → Implementation ○ → Testing ○ → Review ○
+
+────────────────────────────────────────────────────────────
+Summary: 3 active | 5 completed | Avg: 4.2 days/feature
+```
+
+**Legend:**
+- `✓` = Phase completed
+- `●` = Current phase (in progress)
+- `○` = Phase not started
+- `(8/10 tasks)` = Completed tasks / Total tasks
+- `3d` = Days since feature started
+- `Claude` = Last AI that worked on this feature
+
+**If no features exist:**
 ```
 Feature Progress
-════════════════
-user-auth [████████░░] 80%
-  Phase: testing
-  Last: Ran 12 unit tests
+════════════════════════════════════════════════════════════
 
-dashboard [████░░░░░░] 40%
-  Phase: implementation
-  Last: Created 3 components
+No active features.
+
+Start one with: pa:feature "feature name"
 ```
 
 ---
@@ -536,28 +872,888 @@ dashboard [████░░░░░░] 40%
 
 **IMPORTANT:** Show REAL data only, never example data.
 
+**Steps:**
 1. Read `./.proagents/active-features/_index.json`
-2. If empty:
-   ```
-   Project Status
-   ══════════════
-   No features tracked.
-
-   Start with:
-   → pa:feature "name"
-   → pa:fix "bug"
-   ```
-3. If has features:
-   ```
-   Project Status
-   ══════════════
-
-   Active: [count]
-   • [name] - [phase] ([progress]%)
-
-   Paused: [count]
-   Completed: [count]
-   ```
+2. Read current feature's `status.json`
+3. Run: `git status` and `git log --oneline -5`
+4. Read `./.proagents/activity.log | tail -20`
+5. Display enhanced status
+
+**If no features:**
+```
+Project Status
+══════════════════════════════════════════════════════════
+
+No active features.
+
+Start with:
+  → pa:feature "feature name"
+  → pa:fix "bug description"
+```
+
+**If has features (enhanced output):**
+```
+Project Status
+══════════════════════════════════════════════════════════
+
+Feature: user-auth                          Branch: feature/user-auth
+Phase: Implementation (4/6)                 Started: 3d ago
+Progress: ████████████░░░░░░░░ 60%          Time: ~4.5h
+
+Tasks: 3/5 completed
+  ✓ Create auth service
+  ✓ Add login endpoint
+  ✓ Add register endpoint
+  → Implement JWT tokens (in progress)
+  ○ Add password reset
+
+Files: 8 modified (+245, -32)    Tests: 12 ✓ | 2 ✗ | 80% cov
+Contributors: Claude (60%), Cursor (40%)
+
+──────────────────────────────────────────────────────────
+Next: Complete JWT token implementation
+```
+
+**Status indicators:**
+- `✓` = Task completed
+- `→` = In progress
+- `○` = Not started
+- `⚠️ BLOCKED:` = Show if blockers exist
+
+**Summary line (for multiple features):**
+```
+──────────────────────────────────────────────────────────
+Summary: 2 active | 1 paused | 5 completed
+```
+
+---
+
+## pa:activity Execution
+
+**CRITICAL: Show ALL AI platforms, not just your own!**
+
+The purpose of `pa:activity` is cross-AI visibility. Users need to see what ALL AIs have done.
+
+**Command variations:**
+```
+pa:activity                    # All recent activity (default)
+pa:activity --today            # Today's activity only
+pa:activity --ai Claude        # Filter by AI platform
+pa:activity --command pa:fix   # Filter by command type
+pa:activity --files            # Show files changed per action
+```
+
+**Steps:**
+1. Run: `cat .proagents/activity.log | tail -50`
+2. Display ALL entries (Claude, Cursor, ChatGPT, Gemini, Copilot, etc.)
+3. Group by time period (Today, Yesterday, Earlier)
+4. Add action icons based on command type
+5. Show summary with action count per platform
+
+**Action icons:**
+| Command | Icon | Description |
+|---------|------|-------------|
+| `pa:feature` | ✨ | New feature |
+| `pa:fix` | 🐛 | Bug fix |
+| `pa:test` | ✅ | Testing |
+| `pa:doc` | 📝 | Documentation |
+| `pa:implement` | 🔨 | Implementation |
+| `pa:analyze` | 🔍 | Analysis |
+| `pa:design` | 🎨 | Design |
+| `pa:review` | 👀 | Code review |
+| `pa:deploy` | 🚀 | Deployment |
+
+**Output format (enhanced):**
+```
+Recent Activity (All AI Platforms)
+══════════════════════════════════════════════════════════
+
+Today (5 actions)
+─────────────────
+[16:00] [Claude]  ✨ pa:feature - Started user-auth
+        └─ 2 files: feature.json, status.json
+[15:30] [Cursor]  🐛 pa:fix - Fixed login validation bug
+        └─ 1 file: auth.service.ts (+12, -3)
+[15:00] [Gemini]  ✅ pa:test - 12 tests passed, 2 failing
+[14:30] [Claude]  🔨 pa:implement - Created UserService
+        └─ 3 files: user.service.ts, user.controller.ts, user.module.ts
+[14:00] [ChatGPT] 🔍 pa:analyze - Analyzed auth module
+
+Yesterday (8 actions)
+─────────────────────
+[18:00] [Claude]  🎨 pa:design - Created component diagram
+[16:30] [Cursor]  📝 pa:doc - Updated API documentation
+...
+
+──────────────────────────────────────────────────────────
+Summary: 13 actions | 4 AIs active | Last 48h
+  Claude: 5 | Cursor: 4 | Gemini: 2 | ChatGPT: 2
+```
+
+**IMPORTANT:**
+- Show ALL platforms in the log, not just yours
+- This enables cross-AI collaboration
+- Users switch between AIs and need full visibility
+- Group by time for better readability
+- Show files changed when available (from changelog)
+
+---
+
+## pa:deploy Execution
+
+**Show interactive deployment checklist with real-time status.**
+
+**Command variations:**
+```
+pa:deploy                      # Full deployment workflow
+pa:deploy-check                # Pre-deployment checklist only
+pa:deploy-staging              # Deploy to staging
+pa:deploy-prod                 # Deploy to production
+```
+
+**Steps:**
+1. Read `./.proagents/prompts/08-deployment.md`
+2. Run pre-deployment checks (tests, lint, build)
+3. Display interactive checklist with real-time status
+4. Guide through deployment steps
+5. Show post-deployment verification
+
+**Output format (enhanced checklist):**
+```
+Deployment Checklist
+═══════════════════════════════════════════════════════════
+
+📋 Pre-Deployment
+─────────────────
+  ✓ All tests passing                    12/12 tests
+  ✓ No linting errors                    0 errors
+  ✓ Build successful                     2.3s
+  ✓ No console.log statements            cleaned
+  ○ Version bumped                       pending
+  ○ Changelog updated                    pending
+
+🔒 Security
+───────────
+  ✓ No hardcoded secrets                 scanned
+  ✓ Dependencies audited                 0 vulnerabilities
+  ○ Security review approved             awaiting
+
+📦 Build & Assets
+─────────────────
+  ✓ Bundle size acceptable               245kb (limit: 300kb)
+  ✓ Assets optimized                     images compressed
+  ○ Source maps generated                pending
+
+📝 Documentation
+────────────────
+  ✓ README updated                       current
+  ○ API docs updated                     pending
+  ○ Release notes prepared               pending
+
+🚀 Ready to Deploy?
+───────────────────
+Status: 8/14 checks passed
+
+Blockers:
+  ⚠️ Version not bumped - run: npm version patch
+  ⚠️ Changelog needs update - run: pa:changelog
+  ⚠️ Security review pending
+
+Run "pa:deploy --fix" to auto-fix resolvable issues.
+```
+
+**During deployment:**
+```
+Deploying to Staging
+═══════════════════════════════════════════════════════════
+
+[1/5] Building application...        ✓ Complete (2.3s)
+[2/5] Running migrations...          ✓ Complete (0.5s)
+[3/5] Deploying to staging...        ● In progress...
+[4/5] Running smoke tests...         ○ Pending
+[5/5] Verifying health checks...     ○ Pending
+
+Progress: ██████████░░░░░░░░░░ 50%
+```
+
+**Post-deployment verification:**
+```
+Post-Deployment Verification
+═══════════════════════════════════════════════════════════
+
+🏥 Health Checks
+────────────────
+  ✓ API responding                       200 OK (45ms)
+  ✓ Database connected                   healthy
+  ✓ Cache working                        redis OK
+  ✓ External services                    all reachable
+
+📊 Metrics (last 5 min)
+───────────────────────
+  Error rate:     0.1%  (baseline: 0.2%)   ✓ OK
+  Response time:  120ms (baseline: 150ms)  ✓ OK
+  Requests/sec:   450   (baseline: 400)    ✓ OK
+
+🔍 Smoke Tests
+──────────────
+  ✓ Homepage loads                       200ms
+  ✓ Login works                          passed
+  ✓ API endpoints responding             12/12 OK
+  ✓ Critical flows working               all passed
+
+══════════════════════════════════════════════════════════
+✅ Deployment Successful!
+
+Next steps:
+  • Monitor metrics for 30 min
+  • Check error logs: pa:logs --errors
+  • If issues, rollback: pa:rollback
+```
+
+**Rollback available:**
+```
+⚠️ Issues Detected - Rollback Available
+═══════════════════════════════════════════════════════════
+
+Problem: Error rate increased to 5% (threshold: 2%)
+
+Options:
+  [1] Rollback to previous version (v1.2.3)
+  [2] Investigate and fix forward
+  [3] Disable feature flag only
+
+Rollback command: pa:rollback --to v1.2.3
+```
+
+**Legend:**
+- `✓` = Check passed
+- `●` = In progress
+- `○` = Pending
+- `✗` = Failed
+- `⚠️` = Warning/Blocker
+
+---
+
+## pa:test Execution
+
+**Run tests with enhanced coverage visualization.**
+
+**Command variations:**
+```
+pa:test                        # Run all tests with coverage
+pa:test-unit                   # Unit tests only
+pa:test-e2e                    # E2E tests only
+pa:test-watch                  # Watch mode
+pa:test "file"                 # Test specific file
+```
+
+**Steps:**
+1. Read `./.proagents/prompts/06-testing.md`
+2. Detect test framework (Jest, Vitest, Mocha, etc.)
+3. Run tests with coverage
+4. Display enhanced results with visualization
+5. Update `.proagents/worklog/_context.md` with test status
+
+**Output format (enhanced):**
+```
+Test Results
+═══════════════════════════════════════════════════════════
+
+Summary: 45 passed | 2 failed | 3 skipped            2.3s
+
+Coverage
+────────
+Overall:      ████████████████░░░░  78%  (target: 80%)
+
+By Module:
+  src/auth/       ████████████████████  95%  ✓
+  src/api/        ████████████████░░░░  80%  ✓
+  src/services/   ██████████████░░░░░░  72%  ⚠️
+  src/utils/      ██████████░░░░░░░░░░  52%  ✗ below target
+
+Failing Tests
+─────────────
+  ✗ auth.test.ts:45
+    └─ login validation: expected true, got false
+
+  ✗ api.test.ts:102
+    └─ user endpoint: timeout after 5000ms
+
+Slow Tests (>1s)
+────────────────
+  ⚠️ e2e/checkout.test.ts          3.2s
+  ⚠️ integration/api.test.ts       1.8s
+
+═══════════════════════════════════════════════════════════
+Next: Fix failing tests or run "pa:test --update-snapshots"
+```
+
+**If all tests pass:**
+```
+Test Results
+═══════════════════════════════════════════════════════════
+
+✅ All 47 tests passed!                              1.8s
+
+Coverage: ████████████████░░░░ 82% (target: 80%) ✓
+
+═══════════════════════════════════════════════════════════
+Ready for: pa:review or pa:deploy
+```
+
+---
+
+## pa:qa Execution
+
+**Show quality dashboard with all metrics.**
+
+**Command variations:**
+```
+pa:qa                          # Full quality dashboard
+pa:qa-security                 # Security audit only
+pa:qa-performance              # Performance check only
+pa:qa-lint                     # Lint check only
+```
+
+**Steps:**
+1. Run linter (ESLint, etc.)
+2. Run security scan (npm audit, etc.)
+3. Check test coverage
+4. Analyze bundle size
+5. Check documentation coverage
+6. Display unified dashboard
+
+**Output format (quality dashboard):**
+```
+Code Quality Dashboard
+═══════════════════════════════════════════════════════════
+
+Overall Score: ████████████████░░░░ 85% (Good)
+
+🔒 Security
+───────────
+  Status:         ✓ No vulnerabilities
+  Dependencies:   142 packages scanned
+  Secrets:        ✓ No hardcoded secrets found
+  Last audit:     2 hours ago
+
+📏 Linting
+──────────
+  Errors:         0 ✓
+  Warnings:       3 ⚠️
+    └─ unused variable (2), missing return type (1)
+  Auto-fixable:   2
+
+🧪 Test Coverage
+────────────────
+  Coverage:       ████████████████░░░░ 78%
+  Target:         80%
+  Status:         ⚠️ 2% below target
+  Untested:       src/utils/helpers.ts, src/api/legacy.ts
+
+📦 Bundle Size
+──────────────
+  Current:        245kb
+  Limit:          300kb
+  Status:         ✓ Within limit
+  Largest:        react-dom (42kb), lodash (24kb)
+
+📝 Documentation
+────────────────
+  Documented:     ████████████░░░░░░░░ 60%
+  Missing:        15 functions, 3 components
+  README:         ✓ Up to date
+
+🔄 Code Complexity
+──────────────────
+  Average:        8.2 (Good)
+  High (>15):     2 functions
+    └─ processPayment() = 18, validateForm() = 16
+
+═══════════════════════════════════════════════════════════
+Issues: 3 warnings | Suggestions: Run "pa:qa --fix" for auto-fixes
+```
+
+---
+
+## pa:review Execution
+
+**Show code review checklist for current changes.**
+
+**Command variations:**
+```
+pa:review                      # Review current changes
+pa:review --staged             # Review staged files only
+pa:review --pr                 # PR review mode
+```
+
+**Steps:**
+1. Read `./.proagents/prompts/06.5-code-review.md`
+2. Get list of changed files (git diff)
+3. Analyze code for common issues
+4. Display interactive checklist
+5. Highlight issues that need attention
+
+**Output format (review checklist):**
+```
+Code Review Checklist
+═══════════════════════════════════════════════════════════
+
+Files Changed: 8 (+342, -89)
+──────────────────────────────
+
+📋 Code Quality
+───────────────
+  ✓ No console.log statements           cleaned
+  ✓ No debugger statements              cleaned
+  ✓ No commented-out code               cleaned
+  ✓ No TODO without issue number        checked
+  ⚠️ Complex function detected           see below
+
+🔒 Security
+───────────
+  ✓ No hardcoded credentials            scanned
+  ✓ Input validation present            checked
+  ✓ No SQL injection risks              analyzed
+  ○ Auth checks on new endpoints        needs review
+
+🧪 Testing
+──────────
+  ✓ Tests added for new functions       4 new tests
+  ⚠️ Test coverage decreased            -3% (78% → 75%)
+  ○ Edge cases covered                  needs review
+
+📝 Documentation
+────────────────
+  ✓ JSDoc on public functions           present
+  ○ README update needed                new feature
+  ○ API docs update needed              new endpoint
+
+🎨 Style & Conventions
+──────────────────────
+  ✓ Follows naming conventions          checked
+  ✓ Consistent formatting               prettier OK
+  ✓ Import order correct                sorted
+
+⚠️ Attention Required
+─────────────────────
+  src/services/payment.ts:45
+    └─ Cyclomatic complexity: 18 (max: 15)
+    └─ Consider breaking into smaller functions
+
+  src/api/users.ts:102
+    └─ Missing error handling for async call
+
+═══════════════════════════════════════════════════════════
+Status: 12/16 checks passed | 2 warnings | 2 need review
+
+Ready to commit? Address warnings first or run "pa:review --approve"
+```
+
+---
+
+## pa:fix Execution
+
+**Fix bugs with before/after diff and affected tests.**
+
+**Steps:**
+1. Read `./.proagents/workflow-modes/entry-modes.md`
+2. Load context and recent changes
+3. Analyze the bug description
+4. Find and fix the issue
+5. Show before/after diff
+6. Identify affected tests
+7. Run affected tests
+8. Update changelogs
+
+**Output format (bug fix summary):**
+```
+Bug Fix Summary
+═══════════════════════════════════════════════════════════
+
+🐛 Issue: Login validation not checking email format
+
+📁 File: src/auth/login.ts
+
+📝 Changes
+──────────
+Line 45:
+  Before: if (email) {
+  After:  if (email && email.includes('@')) {
+
+Line 52:
+  Before: return { success: true }
+  After:  return { success: true, validated: true }
+
+Diff: +3 lines, -1 line
+
+🧪 Affected Tests
+─────────────────
+  → auth.test.ts (3 tests affected)
+  → validation.test.ts (2 tests affected)
+
+Running affected tests...
+  ✓ auth.test.ts                 5/5 passed
+  ✓ validation.test.ts           4/4 passed
+
+📋 Related
+──────────
+  Issue: #123 (if detected from description)
+  Similar fixes: login.ts (2 weeks ago)
+
+═══════════════════════════════════════════════════════════
+✅ Fix applied and verified!
+
+Changelog updated: .proagents/changelog/_recent.md
+Next: pa:test (full suite) or pa:commit
+```
+
+**If fix causes test failures:**
+```
+Bug Fix Summary
+═══════════════════════════════════════════════════════════
+
+🐛 Issue: Login validation not checking email format
+
+📁 File: src/auth/login.ts
+...
+
+🧪 Affected Tests
+─────────────────
+  ✗ auth.test.ts                 3/5 passed, 2 failed
+    └─ test "allows empty email" now fails (expected behavior?)
+    └─ test "validates email" now fails (needs update)
+
+⚠️ Fix applied but tests failing!
+
+Options:
+  [1] Update tests to match new behavior
+  [2] Revert fix and investigate
+  [3] Keep fix, mark tests as TODO
+
+Run "pa:fix --update-tests" to auto-update test expectations
+```
+
+---
+
+## pa:standup Execution
+
+**Generate daily standup summary automatically.**
+
+**Command variations:**
+```
+pa:standup                     # Today's standup
+pa:standup --yesterday         # What was done yesterday
+pa:standup --week              # Weekly summary
+pa:standup --team              # All AI activity (team view)
+```
+
+**Steps:**
+1. Read `.proagents/activity.log` for recent activity
+2. Read `.proagents/changelog/_recent.md` for changes
+3. Read current feature status from `active-features/`
+4. Identify blockers from context
+5. Generate formatted standup
+
+**Output format:**
+```
+Daily Standup - March 13, 2026
+═══════════════════════════════════════════════════════════
+
+👤 Your Session (Claude)
+────────────────────────
+
+✅ Yesterday / Last Session:
+  • Completed JWT token implementation
+  • Fixed 2 login validation bugs (#123, #124)
+  • Added 8 unit tests for auth module
+  • Updated API documentation
+
+📋 Today / Current:
+  • Password reset feature (in progress - 40%)
+    └─ Next: Email template integration
+  • API rate limiting implementation
+  • Review PR #45 (user dashboard)
+
+🚧 Blockers:
+  • None currently
+
+📊 Stats:
+  Files changed: 12 | Tests: +8 | Coverage: 78% → 82%
+
+──────────────────────────────────────────────────────────
+
+👥 Team Activity (All AIs) - Last 24h
+─────────────────────────────────────
+  Claude:   5 tasks  │ auth module, API docs
+  Cursor:   3 tasks  │ UI components, styling
+  Gemini:   2 tasks  │ database optimization
+
+═══════════════════════════════════════════════════════════
+Active Features: user-auth (80%), dashboard (40%)
+```
+
+**Weekly summary (pa:standup --week):**
+```
+Weekly Summary - Week of March 10, 2026
+═══════════════════════════════════════════════════════════
+
+📈 Progress
+───────────
+  Features completed:    2
+  Features in progress:  3
+  Bugs fixed:           8
+  Tests added:          24
+
+📊 By AI Platform
+─────────────────
+  Claude    ████████████████░░░░  45%  (18 tasks)
+  Cursor    ████████████░░░░░░░░  32%  (13 tasks)
+  Gemini    ████████░░░░░░░░░░░░  23%  (9 tasks)
+
+✅ Completed
+────────────
+  • User authentication feature
+  • Dashboard redesign
+  • 8 bug fixes
+
+🚀 Shipped
+──────────
+  • v1.2.0 deployed to production (March 12)
+
+📋 Next Week
+────────────
+  • Password reset feature
+  • Payment integration
+  • Performance optimization
+
+═══════════════════════════════════════════════════════════
+Total: 40 tasks | 156 files changed | Coverage: 75% → 82%
+```
+
+---
+
+## Self-Learning System
+
+**ProAgents learns from user interactions. AI MUST read and write learning data.**
+
+### Learning Data Location
+
+```
+.proagents/.learning/
+├── global/                      # Cross-project preferences
+│   ├── user-preferences.json    # Checkpoint & workflow preferences
+│   └── common-patterns.json     # Common code patterns
+│
+└── projects/                    # Per-project learnings
+    └── {project-id}/
+        ├── patterns.json        # Project-specific code patterns
+        ├── corrections.json     # User corrections to AI output
+        ├── metrics.json         # Performance metrics
+        └── feedback.json        # User feedback
+```
+
+### On Session Start - Load Learning Data
+
+**AI MUST read learning data at start of session:**
+
+```bash
+# Load user preferences
+cat .proagents/.learning/global/user-preferences.json 2>/dev/null
+
+# Load project patterns
+cat .proagents/.learning/projects/*/patterns.json 2>/dev/null
+
+# Load past corrections (DON'T REPEAT THESE MISTAKES!)
+cat .proagents/.learning/projects/*/corrections.json 2>/dev/null
+```
+
+### Apply Learned Patterns
+
+**Before generating code, check learned patterns:**
+
+```json
+// From user-preferences.json
+{
+  "code_preferences": {
+    "naming_style": { "components": "PascalCase" },
+    "patterns": { "state_management": "zustand" }
+  }
+}
+```
+
+**AI should say:**
+```
+"I see this project uses PascalCase for components and Zustand for state.
+ I'll follow those patterns."
+```
+
+### When to Write Learning Data
+
+**1. User Corrects AI Output**
+
+When user modifies AI-generated code, log to `corrections.json`:
+
+```json
+{
+  "corrections": [
+    {
+      "timestamp": "2024-03-13T10:30:00Z",
+      "ai": "Claude",
+      "type": "naming",
+      "original": "getData",
+      "corrected": "fetchUserData",
+      "context": "User prefers descriptive function names",
+      "occurrences": 1
+    }
+  ]
+}
+```
+
+**2. User Skips/Reviews Checkpoints**
+
+Track checkpoint preferences in `user-preferences.json`:
+
+```json
+{
+  "checkpoint_preferences": {
+    "usually_skips": ["after_requirements"],
+    "always_reviews": ["before_deployment"]
+  }
+}
+```
+
+**3. User Provides Explicit Feedback**
+
+When user says "that's wrong" or corrects approach, log to `feedback.json`:
+
+```json
+{
+  "feedback": [
+    {
+      "timestamp": "2024-03-13T10:30:00Z",
+      "type": "correction",
+      "context": "API call pattern",
+      "feedback": "Always use try-catch for API calls",
+      "applied": true
+    }
+  ]
+}
+```
+
+### pa:learn Command
+
+**User can explicitly teach AI a pattern:**
+
+```
+pa:learn "Always use async/await instead of .then()"
+pa:learn "Components go in src/components/, not src/ui/"
+pa:learn "Use date-fns, not moment.js"
+```
+
+**AI logs to patterns.json:**
+```json
+{
+  "learned_patterns": [
+    {
+      "timestamp": "2024-03-13T10:30:00Z",
+      "pattern": "Always use async/await instead of .then()",
+      "source": "explicit",
+      "confidence": 1.0
+    }
+  ]
+}
+```
+
+### pa:forget Command
+
+**User can remove a learned pattern:**
+
+```
+pa:forget "Use moment.js for dates"
+```
+
+**AI removes from patterns.json and confirms.**
+
+### Auto-Apply Rules
+
+| Correction Count | Action |
+|------------------|--------|
+| 1-2 times | Remember but don't auto-apply |
+| 3-4 times | Warn: "Should I always do this?" |
+| 5+ times | Auto-apply without asking |
+
+### Example: Learning in Action
+
+**Session 1:**
+```
+User: Fix the login function
+AI: [generates code with getData()]
+User: [changes getData() to fetchUserData()]
+AI: [logs correction to corrections.json]
+```
+
+**Session 2 (same or different AI):**
+```
+AI: [reads corrections.json]
+AI: "I see you prefer descriptive names like fetchUserData.
+     I'll use that pattern."
+AI: [generates code with fetchUserData()]
+```
+
+### Learning Report (pa:learning)
+
+```
+pa:learning
+```
+
+**Output:**
+```
+Learning Report
+═══════════════════════════════════════════════════════════
+
+📊 Summary
+──────────
+  Patterns learned:     12
+  Corrections logged:   8
+  Auto-apply rules:     3
+
+🎯 Top Patterns
+───────────────
+  • Naming: Use descriptive function names (5 corrections)
+  • State: Use Zustand for state management
+  • API: Always use try-catch for API calls
+  • Testing: Use describe/it pattern for tests
+
+⚠️ Recent Corrections
+─────────────────────
+  • getData → fetchUserData (3 times)
+  • Component → src/components/ not src/ui/
+
+✅ Auto-Apply Rules (5+ corrections)
+────────────────────────────────────
+  • Use PascalCase for components
+  • Use camelCase for functions
+  • Add error handling to API calls
+
+═══════════════════════════════════════════════════════════
+```
+
+### CRITICAL: Don't Repeat Mistakes
+
+**Before generating code, AI MUST check:**
+
+1. `corrections.json` - Don't make same mistakes
+2. `feedback.json` - Follow user feedback
+3. `patterns.json` - Use learned patterns
+
+**Example check:**
+```bash
+# Check if similar code was corrected before
+grep -i "getData" .proagents/.learning/projects/*/corrections.json
+```
+
+If found, apply the correction pattern automatically.
 
 ---
 
@@ -586,6 +1782,9 @@ Or ask user: "Should I read the detailed docs for [topic]?"
 | `watchlist.yaml` | Files requiring confirmation |
 | `proagents.config.yaml` | Project config |
 | `active-features/_index.json` | Feature status |
+| `.learning/global/user-preferences.json` | User workflow preferences |
+| `.learning/projects/*/patterns.json` | Learned code patterns |
+| `.learning/projects/*/corrections.json` | Past corrections (DON'T REPEAT!) |
 
 ---