projecta-rrr 1.16.5 → 1.16.7

package/CHANGELOG.md

@@ -4,7 +4,56 @@ All notable changes to RRR will be documented in this file.
 
 Format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
-## [1.16.5] - 2026-01-27
+## [1.16.7] - 2026-01-28
+
+### Added
+- **Memory System** (`rrr/lib/memory-store.js`) - Central library for tracking:
+  - Command history with intent extraction
+  - Current user intent persistence
+  - Decision log with rationale
+  - Drift detection (context drift, stale plans, decision conflicts)
+
+- **Visual HUD** (`scripts/rrr-hud.js`) - ASCII progress display showing:
+  - Milestone progress bar with percentage
+  - Per-phase completion status
+  - Drift status (green/yellow/red)
+  - Current intent from memory
+
+- **Proactive Guidance** (`scripts/rrr-memory/state-detector.js`) - State-aware suggestions:
+  - 4D state analysis (project, milestone, phase, action)
+  - `guidanceFor(command)` for 17 command scenarios
+  - Orphan phase detection and migration guidance
+
+- **Memory-Enriched Preflight** (`scripts/handoff-preflight.js`) - Drift detection on handoff
+
+### Changed
+- **Phase path resolution** (`rrr/lib/phase-paths.md`) - Fixed bug where phases were created outside milestones when no `:construction:` marker existed. Now errors if `.planning/milestones/` exists but no active milestone detected.
+
+- **execute-plan** - Tracks intent before execution for drift detection
+
+- **verify-work** - Shows memory context during verification
+
+- **add-phase / insert-phase** - Validates no orphan phases exist before allowing new phase creation
+
+- **Duplicate prevention** (`bin/install.js`, `scripts/doctor-rrr.js`) - Auto-cleans old quarantine directories (>7 days) to prevent duplicate command scanning
+
+### Fixed
+- **Duplicate commands** - 282 command references reduced to 41 by cleaning old quarantines
+
+## [1.16.6] - 2026-01-28
+
+### Added
+- **Proactive Progress Routing** - Enhanced `/rrr:progress` to ask clarifying questions:
+  - **Route C** (phase complete): Asks "Continue to next phase?", "Discuss first?", or "Different work?"
+  - **Route F.1** (between milestones with context): Asks "Create milestone now?", "Discuss more?", or "Quick fix?"
+  - **Route F.2** (between milestones no context): Asks about work scope (quick fix, new feature, new milestone)
+  - Routes based on user intent, not just sequential progression
+
+### Changed
+- **Progress command now uses AskUserQuestion** for all decision points instead of just presenting options
+- Clearer guidance for Patch vs Phase vs Milestone decisions at milestone boundaries
+
+## [1.16.5] - 2026-01-28
 
 ### Fixed
 - **Verify-Work UAT Workflow Path** - Fixed workflow file loading for UAT mode:
@@ -17,9 +66,57 @@ Format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
   - Fixed commit message generation for milestone-level UAT
 - **PHASE_ARG Variable Propagation** - Added export of `PHASE_ARG` for workflow consumption
 - **SUMMARY Path Detection** - Fixed test type detection to search both old and new directory structures
-- **UAT File Discovery** - Updated active session detection to search `.planning/` recursively for both phase and milestone UAT files
+- **UAT File Discovery** - Updated active session detection to search `.planning/` recursively
+
+### Added
+- **Project State Decision Tree** - New reference document at `rrr/references/project-state-flow.md`:
+  - Visual state machine diagram showing all project states and transitions
+  - Decision tables for each state (Between Milestones, Active Milestone, Active Phase, Phase Complete, Milestone Complete)
+  - Quick reference: which command to use for each scenario
+  - Patch vs Phase vs Milestone comparison guide
+  - Decimal phase semantics explained
+
+### Changed
+- **Proactive Progress Routing** - Enhanced `/rrr:progress` command to ask clarifying questions:
+  - **Route C** (phase complete): Asks "Continue to next phase?", "Discuss first?", or "Different work?"
+  - **Route F.1** (between milestones with context): Asks "Create milestone now?", "Discuss more?", or "Quick fix?"
+  - **Route F.2** (between milestones no context): Asks about work scope (quick fix, new feature, new initiative)
+  - Clear routing based on user intent rather than assuming sequential progression
 
 ## [1.16.4] - 2026-01-27
+
+### Fixed
+- **Playwright MCP Detection** - Enhanced detection logic now checks multiple locations:
+  - `~/.claude/mcp.json` (primary)
+  - `~/.claude/settings.json` (alternate)
+  - `playwright-mcp` command availability
+  - Running processes via `ps`
+- **QA Engineer Mode Auto-Detection** - UAT now auto-detects test type from SUMMARY.md:
+  - Web UI tests → Recommend AI QA Engineer (Playwright + MiniMax)
+  - API/Backend tests → Recommend Zeroshot
+  - Mixed tests → Offer both options with clear recommendations
+
+### Added
+- **Intuitive Mode Selection** - Testing mode selection now shows:
+  - Auto-detected test type (e.g., "Web UI tests")
+  - Clear explanation of each mode's strengths
+  - Recommended option highlighted
+  - Cost estimates for AI-powered modes
+
+- **run_qa_engineer_batch Step** - New workflow step for AI QA Engineer mode:
+  - Combines Playwright MCP (browser control) + MiniMax (visual AI)
+  - Tests edge cases automatically
+  - Explores navigation paths beyond happy path
+  - ~$0.001 per test (MiniMax vision is 50-100x cheaper than Anthropic API)
+
+- **run_playwright_batch Step** - New workflow step for Playwright-only mode:
+  - Browser automation without visual AI validation
+  - Fallback when MiniMax API key not available
+
+### Changed
+- **Improved Mode Routing** - create_uat_file step now routes to 4 modes:
+  - `qa-engineer` → run_qa_engineer_batch
+  - `zeroshot` → run_zeroshot_batch
   - `playwright` → run_playwright_batch
   - `manual` → present_test
 

package/bin/install.js

@@ -11,7 +11,8 @@ const {
   detectAndQuarantineDuplicates,
   writeInstallInfo,
   getCurrentVersion,
-  getCanonicalRoots
+  getCanonicalRoots,
+  cleanOldQuarantines
 } = require('../rrr/lib/install-roots');
 
 /**
@@ -909,6 +910,38 @@ function install(isGlobal) {
     console.log(`  ${green}✓${reset} Installed rrr/scripts/test-install-smoke.js`);
   }
 
+  // PATCH-01: Copy handoff-preflight.js to ~/.claude/rrr/scripts/
+  // Memory-enriched preflight for drift detection
+  const handoffSrc = path.join(src, 'scripts', 'handoff-preflight.js');
+  if (fs.existsSync(handoffSrc)) {
+    const scriptsDestDir = path.join(claudeDir, 'rrr', 'scripts');
+    fs.mkdirSync(scriptsDestDir, { recursive: true });
+    const handoffDest = path.join(scriptsDestDir, 'handoff-preflight.js');
+    fs.copyFileSync(handoffSrc, handoffDest);
+    console.log(`  ${green}✓${reset} Installed rrr/scripts/handoff-preflight.js`);
+  }
+
+  // PATCH-01: Copy rrr-hud.js to ~/.claude/rrr/scripts/
+  // Visual HUD for project state display
+  const hudSrc = path.join(src, 'scripts', 'rrr-hud.js');
+  if (fs.existsSync(hudSrc)) {
+    const scriptsDestDir = path.join(claudeDir, 'rrr', 'scripts');
+    fs.mkdirSync(scriptsDestDir, { recursive: true });
+    const hudDest = path.join(scriptsDestDir, 'rrr-hud.js');
+    fs.copyFileSync(hudSrc, hudDest);
+    console.log(`  ${green}✓${reset} Installed rrr/scripts/rrr-hud.js`);
+  }
+
+  // PATCH-01: Copy rrr-memory/ directory to ~/.claude/rrr/scripts/
+  // Memory store and state detector for proactive guidance
+  const rrrMemorySrc = path.join(src, 'scripts', 'rrr-memory');
+  if (fs.existsSync(rrrMemorySrc)) {
+    const scriptsDestDir = path.join(claudeDir, 'rrr', 'scripts');
+    const rrrMemoryDest = path.join(scriptsDestDir, 'rrr-memory');
+    copyWithPathReplacement(rrrMemorySrc, rrrMemoryDest, pathPrefix);
+    console.log(`  ${green}✓${reset} Installed rrr/scripts/rrr-memory/`);
+  }
+
   // Copy skills to ~/.claude/skills (skills system)
   const skillsSrc = path.join(src, 'rrr', 'skills');
   if (fs.existsSync(skillsSrc)) {
@@ -963,6 +996,15 @@ function install(isGlobal) {
     if (quarantineResult.quarantined.length > 0) {
       console.log(`  ${green}✓${reset} Quarantined ${quarantineResult.quarantined.length} legacy/duplicate root(s)`);
     }
+
+    // Clean up old quarantine directories (>7 days) to prevent duplicate command scanning
+    console.log(`  ${green}✓${reset} Cleaning old quarantine directories...`);
+    const cleanupResult = cleanOldQuarantines({
+      configDir: explicitConfigDir || process.env.CLAUDE_CONFIG_DIR
+    });
+    if (cleanupResult.cleaned.length > 0) {
+      console.log(`  ${green}✓${reset} Cleaned ${cleanupResult.cleaned.length} old quarantine folder(s)`);
+    }
   }
 
   // Register rrr-search MCP server and run semantic migration

package/commands/rrr/add-phase.md

@@ -33,6 +33,7 @@ Purpose: Add planned work discovered during execution that belongs at the end of
 @.planning/STATE.md
 @rrr/lib/phase-paths.md
 @rrr/lib/milestone-utils.md
+@rrr/lib/memory-store.js
 </execution_context>
 
 <process>
@@ -68,6 +69,50 @@ Example: /rrr:add-phase Add authentication system
 Exit.
 </step>
 
+<step name="check_orphan_phases_memory">
+**Check for orphan phases using state-detector (PATCH-01 enhancement)**
+
+Before creating phases, check if there are orphan phases that need migration:
+
+```bash
+# Check for orphan phases
+node -e "
+const { StateDetector } = require(process.env.HOME + '/.claude/rrr/scripts/rrr-memory/state-detector');
+const detector = new StateDetector();
+const state = detector.detectPhaseState('.planning');
+const milestone = detector.detectMilestoneState('.planning');
+
+if (milestone === 'NO_MILESTONE' && state !== 'NO_PHASES') {
+  console.log('HAS_ORPHAN_PHASES=true');
+  console.log('RUN: /rrr:new-milestone or /rrr:migrate-phases');
+}
+" 2>/dev/null
+```
+
+If `HAS_ORPHAN_PHASES=true` is output, show guidance and exit (unless --force).
+
+**If orphan phases detected:**
+
+```
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+⚠ ORPHAN PHASES DETECTED
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+Found phases not assigned to any milestone.
+New phases cannot be created until orphans are assigned.
+
+To migrate orphan phases:
+1. /rrr:new-milestone - Create milestone for orphan phases
+2. /rrr:migrate-phases - Assign phases to milestone
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+Exit command.
+
+**If no orphans or --force:** Continue to validate_milestone_first step.
+</step>
+
 <step name="validate_milestone_first">
 **MILESTONE-FIRST ENFORCEMENT (v1.11 breaking change)**
 

package/commands/rrr/execute-plan.md

@@ -36,6 +36,7 @@ Plan path: $ARGUMENTS
 
 @.planning/STATE.md
 @.planning/config.json (if exists)
+@rrr/lib/memory-store.js
 </context>
 
 <process>
@@ -46,6 +47,42 @@ Plan path: $ARGUMENTS
    ```
    Cross-platform: works in PowerShell on Windows. Path resolved to installed location.
 
+0. **Track command in memory (before execution)**
+   Parse the plan to extract intent from frontmatter and context, then track this execution:
+
+   ```bash
+   # Extract plan ID and objective for intent tracking
+   PLAN_ID=$(basename "$ARGUMENTS" | sed 's/-PLAN\.md$//')
+   PLAN_DIR=$(dirname "$ARGUMENTS")
+   PHASE_DIR=$(basename "$PLAN_DIR")
+
+   # Get intent from plan frontmatter (try files_modified, then title)
+   INTENT=""
+   if [ -f "$ARGUMENTS" ]; then
+     INTENT=$(grep -E "^files_modified:" "$ARGUMENTS" 2>/dev/null | head -1 | sed 's/files_modified: *\[*\]*/Intent from plan/' || \
+              grep -E "^title:" "$ARGUMENTS" 2>/dev/null | head -1 | sed 's/title: *//' || \
+              echo "Executing plan $PLAN_ID")
+   fi
+
+   # Track in memory (if memory-store.js is available)
+   node -e "
+   const { initMemory, trackCommand } = require(process.env.HOME + '/.claude/rrr/lib/memory-store');
+   (async () => {
+     try {
+       const memory = await initMemory();
+       await trackCommand(memory, {
+         command: '/rrr:execute-plan',
+         target: '$PLAN_ID',
+         intent: '$INTENT',
+         files: []
+       });
+     } catch (e) {
+       // Memory tracking is optional
+     }
+   })();
+   " 2>/dev/null || echo "[MEMORY] Tracking skipped"
+   ```
+
 0.5. **Scope Check (validate plan is in scope)**
 
    Before executing, validate the plan's files against current scope:

package/commands/rrr/insert-phase.md

@@ -23,6 +23,7 @@ Purpose: Handle urgent work discovered during execution without renumbering enti
 @.planning/STATE.md
 @rrr/lib/phase-paths.md
 @rrr/lib/milestone-utils.md
+@rrr/lib/memory-store.js
 </execution_context>
 
 <process>
@@ -74,6 +75,50 @@ fi
 
 </step>
 
+<step name="check_orphan_phases_memory">
+**Check for orphan phases using state-detector (PATCH-01 enhancement)**
+
+Same validation as add-phase - check for orphan phases before allowing new insert:
+
+```bash
+# Check for orphan phases
+node -e "
+const { StateDetector } = require(process.env.HOME + '/.claude/rrr/scripts/rrr-memory/state-detector');
+const detector = new StateDetector();
+const state = detector.detectPhaseState('.planning');
+const milestone = detector.detectMilestoneState('.planning');
+
+if (milestone === 'NO_MILESTONE' && state !== 'NO_PHASES') {
+  console.log('HAS_ORPHAN_PHASES=true');
+  console.log('RUN: /rrr:new-milestone or /rrr:migrate-phases');
+}
+" 2>/dev/null
+```
+
+If `HAS_ORPHAN_PHASES=true`, show guidance and exit (unless --force).
+
+**If orphan phases detected:**
+
+```
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+⚠ ORPHAN PHASES DETECTED
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+
+Found phases not assigned to any milestone.
+Inserting new phase not allowed until orphans are assigned.
+
+To migrate orphan phases:
+1. /rrr:new-milestone - Create milestone for orphan phases
+2. /rrr:migrate-phases - Assign phases to milestone
+
+━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
+```
+
+Exit command.
+
+**If no orphans or --force:** Continue to validate_milestone_first step.
+</step>
+
 <step name="validate_milestone_first">
 **MILESTONE-FIRST ENFORCEMENT (v1.11 breaking change)**
 

package/commands/rrr/progress.md

@@ -461,20 +461,60 @@ Include audit summary in output (if run).
 
 Full report: .planning/artifacts/VERIFY_WORK_AUDIT.md
 
-## ▶ Next Up
+<AskUserQuestion>
+<questions>
+[
+  {
+    "question": "Phase {Z} is complete. What's next?",
+    "header": "Next",
+    "options": [
+      {
+        "label": "Continue to Phase {Z+1}",
+        "description": "Execute next phase in current milestone"
+      },
+      {
+        "label": "Discuss Phase {Z+1} first",
+        "description": "Gather context before planning (recommended for UI/UX)"
+      },
+      {
+        "label": "Actually, different work",
+        "description": "Maybe a patch, or skip to another phase?"
+      }
+    ]
+  }
+]
+</questions>
+</AskUserQuestion>
 
-**Phase {Z+1}: {Name}** — {Goal from ROADMAP.md}
+---
 
-`/rrr:plan-phase {Z+1}`
+**Based on response:**
 
-<sub>`/clear` first → fresh context window</sub>
+| Response | Action | Command |
+|----------|--------|---------|
+| Continue to Phase {Z+1} | Plan and execute | `/rrr:plan-phase {Z+1}` |
+| Discuss first | Gather context | `/rrr:discuss-phase {Z+1}` |
+| Different work | Clarify intent | Present options (patch, skip phase, complete milestone) |
+
+**For "Actually, different work" response:**
+```
+## → What would you like to work on instead?
+
+**Options:**
+
+1. **Quick fix/patch** — `/rrr:insert-phase` as decimal (e.g., {Z}.1)
+2. **Skip to different phase** — `/rrr:execute-phase {XX}`
+3. **Complete milestone** — All remaining phases done? `/rrr:complete-milestone`
+4. **Different scope entirely** — Between milestones? `/rrr:discuss-milestone`
+
+What type of work did you have in mind?
+```
 
 ---
 
 **Also available:**
 - `/rrr:verify-work --uat {Z}` — user acceptance test before continuing
 - `/rrr:verify-work --audit` — full implementation audit
-- `/rrr:discuss-phase {Z+1}` — gather context first
 
 ---
 ```
@@ -560,19 +600,40 @@ Key features from discussion:
 - [feature 2]
 - [feature 3]
 
-## ▶ Next Up
-
-**Create the milestone** — formalize what you discussed
-
-`/rrr:new-milestone`
-
-<sub>`/clear` first → fresh context window</sub>
+<AskUserQuestion>
+<questions>
+[
+  {
+    "question": "Ready to formalize this milestone, or want to discuss more?",
+    "header": "Milestone",
+    "options": [
+      {
+        "label": "Create milestone now",
+        "description": "Formalize vX.(Y+1) with requirements and roadmap"
+      },
+      {
+        "label": "Discuss more first",
+        "description": "Continue the discussion, add more context"
+      },
+      {
+        "label": "Actually, it's a quick fix",
+        "description": "This is just a patch, not a new milestone"
+      }
+    ]
+  }
+]
+</questions>
+</AskUserQuestion>
 
 ---
 
-**Also available:**
-- `/rrr:discuss-milestone` — continue the discussion
-- `/rrr:verify-work --audit` — audit completed work
+**Based on response:**
+
+| Response | Action | Command |
+|----------|--------|---------|
+| Create milestone now | Formalize vX.(Y+1) | `/rrr:new-milestone` |
+| Discuss more | Continue discussion | `/rrr:discuss-milestone` |
+| Quick fix | Create patch | `/rrr:insert-phase` (decimal) |
 
 ---
 ```
@@ -612,29 +673,99 @@ Full report: .planning/artifacts/VERIFY_WORK_AUDIT.md
 
 ---
 
-## What's Next?
+## What would you like to do next?
 
-**Choose based on scope of work:**
+<AskUserQuestion>
+<questions>
+[
+  {
+    "question": "What's the scope of your next piece of work?",
+    "header": "Scope",
+    "options": [
+      {
+        "label": "Quick fix/hotfix",
+        "description": "Bug fix, small tweak, or one-line change → Creates a PATCH (vX.Y.1)"
+      },
+      {
+        "label": "New feature (current milestone)",
+        "description": "Fits current milestone scope → Creates a new PHASE in vX.Y"
+      },
+      {
+        "label": "New initiative (new milestone)",
+        "description": "New scope, capabilities, or requirements → Creates MILESTONE vX.(Y+1)"
+      }
+    ]
+  }
+]
+</questions>
+</AskUserQuestion>
 
-### Small Work → Track as Patch
-Bugs, hotfixes, small improvements belong in v{X.Y} as patches.
-- Document in `.planning/milestones/v{X.Y}/PATCHES.md`
-- Commit with `fix:` or `chore:` prefix
-- Bump patch version (v{X.Y}.1, v{X.Y}.2, etc.)
-- No formal phase/plan needed
+---
 
-### Larger Scope → Start New Milestone
-Features, multi-file changes, new capabilities need proper planning.
+**Based on response, route accordingly:**
 
----
+| Response | Action | Command |
+|----------|--------|---------|
+| Quick fix/hotfix | Create decimal phase as patch | `/rrr:insert-phase` (decimal like 9.1) |
+| New feature | Add phase to current milestone | `/rrr:add-phase` |
+| New initiative | Discuss and create milestone | `/rrr:discuss-milestone` |
 
-## ▶ Next Up (for new milestone)
+**For "Quick fix/hotfix" response:**
+```
+## → Creating a Patch (v{X.Y}.1)
 
-**Discuss what to build next** — thinking partner for next milestone
+Quick fixes don't need full phase/plan structure.
 
-`/rrr:discuss-milestone`
+**To create a patch:**
+1. `/rrr:insert-phase` — I'll insert as decimal phase (e.g., {current_phase}.1)
+2. This creates a minimal phase directory for tracking
+3. Or simply commit directly with `fix:` prefix — patches are documented in PATCHES.md
 
-<sub>`/clear` first → fresh context window</sub>
+**Patch vs Phase Decision Guide:**
+- Patch: Hours of work, 1-2 files changed, bug fix
+- Phase: Days of work, multiple files, feature addition
+
+Want me to create a decimal phase for this patch?
+```
+
+**For "New feature (current milestone)" response:**
+```
+## → Adding Phase to v{X.Y}
+
+**Current phases in v{X.Y}:**
+- [list phases from ROADMAP.md]
+
+**To add a new phase:**
+`/rrr:add-phase` — I'll ask for phase number and description
+
+**Phase Naming:**
+- Next sequential: {next_phase_num}
+- Or specify custom number
+
+Ready to add the phase?
+```
+
+**For "New initiative (new milestone)" response:**
+```
+## → Starting New Milestone v{X.(Y+1)}
+
+**To begin a new milestone:**
+
+1. **Discuss first** (recommended): `/rrr:discuss-milestone`
+   - Thinking partner for next scope
+   - Creates MILESTONE-CONTEXT.md with decisions
+   - Then `/rrr:new-milestone` formalizes it
+
+2. **Skip discussion**: `/rrr:new-milestone`
+   - Directly creates milestone structure
+   - Less context gathered upfront
+
+**New Milestone vs Patch Decision Guide:**
+- New Milestone: New capabilities, integrations, multi-sprint work
+- Patch: Quick fixes, small improvements, bug fixes
+
+How would you like to proceed?
+```
 
 ---
 

package/commands/rrr/verify-work.md

@@ -71,6 +71,7 @@ Validate built features through **audit mode by default**, or interactive UAT wi
 
 <execution_context>
 @rrr/lib/phase-paths.md
+@rrr/lib/memory-store.js
 </execution_context>
 
 <!-- ═══════════════════════════════════════════════════════════════════════════
@@ -140,6 +141,41 @@ For manual testing, use: /rrr:verify-work --uat [phase]
 
 ---
 
+### Step 0.5: Show Memory Context (PATCH-01 enhancement)
+
+Display current intent and recent commands to provide context during verification:
+
+```bash
+# Show memory context if available
+node -e "
+const { initMemory, getSummary } = require(process.env.HOME + '/.claude/rrr/lib/memory-store');
+(async () => {
+  try {
+    const memory = await initMemory();
+    const summary = await getSummary(memory);
+    console.log('');
+    console.log('━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━');
+    console.log(' RRR ► MEMORY CONTEXT');
+    console.log('━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━');
+    console.log('Intent:', summary.currentIntent || '(no intent tracked)');
+    console.log('Recent:', summary.recentCommands.slice(0, 2).map(c =>
+      c.cmd + (c.target ? '(' + c.target + ')' : '')
+    ).join(' → ') || 'none');
+    if (summary.drift && summary.drift.drifting) {
+      console.log('Drift:', summary.drift.message);
+    }
+    console.log('───────────────────────────────────────────────────────');
+  } catch (e) {
+    // Memory not available, skip silently
+  }
+})();
+" 2>/dev/null || true
+```
+
+This helps verify-work understand what the user was working on.
+
+---
+
 ### Step 1: Determine Scope
 
 **Parse target from arguments (with flags removed):**

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "projecta-rrr",
-  "version": "1.16.5",
+  "version": "1.16.7",
   "description": "A meta-prompting, context engineering and spec-driven development system for Claude Code by Projecta.ai",
   "bin": {
     "projecta-rrr": "bin/install.js"
@@ -22,7 +22,10 @@
     "sessions:index": "bash scripts/index-sessions.sh",
     "sessions:status": "bash scripts/index-sessions.sh --status",
     "semantic:verify": "bash scripts/verify-semantic-search.sh",
-    "prepublish:check": "node scripts/prepublish-check.js"
+    "prepublish:check": "node scripts/prepublish-check.js",
+    "browser:uat": "bash scripts/browser-uat.sh",
+    "scaffold:e2e": "bash scripts/scaffold-e2e.sh",
+    "chrome:visual": "bash scripts/chrome-visual-check.sh"
   },
   "files": [
     "bin",