mover-os 4.0.13 → 4.0.15

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mover-os",
-  "version": "4.0.13",
+  "version": "4.0.15",
   "description": "The self-improving OS for AI agents. Turns Obsidian into an execution engine.",
   "bin": {
     "mover-os": "install.js"

package/src/system/Mover_Global_Rules.md

@@ -324,6 +324,7 @@ These triggers fire during EVERY conversation, not just workflows. **DETECT > WR
 | Idea killed | "We're not doing X", "scrap that", "dead idea" | Active_Context `## 2. THE GRAVEYARD` |
 | URL/content pasted | Link to article, video, tweet, reel | Ask: "Capture to Library/Inputs?" then write on confirmation |
 | Named person mentioned | "I spoke to [Name]" (new, not in Entities/) | Ask: "Create entity note for [Name]?" |
+| User corrects AI behavior | "Don't do that", "Why did you do X?", "From now on do Y" | Execute S11 Trigger A inline — propose fix, wait for approval |
 
 The file is the truth, not the conversation. Chat-only decisions evaporate between sessions.
 
@@ -397,14 +398,47 @@ The file is the truth, not the conversation. Chat-only decisions evaporate betwe
 
 ## 11. Self-Improvement & Pattern Detection
 
-**Trigger A: AI Correction** — User corrects AI behavior.
+**Trigger A: AI Correction** — User corrects AI behavior ("Why did you do that?", "Don't do X", "From now on do Y").
+
 1. Acknowledge: "I made a mistake."
-2. Update the relevant file so the next instance won't repeat it.
-3. Ask user to review the update.
-4. Log correction in `Auto_Learnings.md` with: Date, Error, Correction, Confidence: 3.
-5. If same error caught 3x, confidence increases to 5.
+2. Identify the violated rule AND the source file (workflow or Global Rules).
+3. Propose a specific fix using diff format:
+   ```
+   CORRECTION FIX:
+   - File: [path]
+   - OLD: "[current text]"
+   + NEW: "[proposed text]"
+   SAFETY CHECK: [Which Safe Modification gate this passes]
+   ```
+4. **STOP. Wait for user approval.**
+5. If approved: implement immediately + log structured CORRECTION to Auto_Learnings.md:
+   ```
+   ## CORRECTION: [Error Name]
+   **Domain:** [AI Behavior / Workflow / Rule]
+   **Error:** [What the AI did wrong]
+   **Fix:** [What the correct behavior is]
+   **Status:** ACTIVE
+   **First Detected:** YYYY-MM-DD
+   **Confidence:** 3
+   **Last Updated:** YYYY-MM-DD
+   ### Evidence
+   - YYYY-MM-DD: "[description]"
+   ```
+6. If deferred: log with `Status: DEFERRED` — review-week 3H picks it up.
+7. If same correction 3x+: confidence escalates to 5, flag as structural.
 - NEVER argue or justify. Just fix and learn.
 
+**Safe System Modification Protocol (mandatory for all self-edits):**
+When the system proposes edits to its own workflow or rule files, these guardrails apply:
+
+1. **Scope Constraint** — Only edit the specific line that caused the issue. Never rewrite surrounding logic.
+2. **Leverage Check (mandatory)** — Same 4 gates from review-week 3H: (a) Does the model already handle this? (b) Enforcement or guidance? (c) Global or workflow-level? (d) Duplicate?
+3. **Diff Format (mandatory)** — Always show OLD and NEW. Never describe the change — show it.
+4. **User Gate (mandatory)** — STOP after showing diff. No auto-edits to workflow/rule files ever.
+5. **Blast Radius Check** — If change affects >1 file: "Recommend deferring to /review-week 3H for batch handling."
+6. **Rollback Path** — Log every approved edit to Active_Context `## My Customizations` table.
+7. **Excluded workflows** — Never self-edit during /setup, /walkthrough, /update (these have their own modification logic).
+
 **Trigger B: User Behavioral Pattern** — AI detects recurring behavior in Daily Notes/Session Logs.
 - Examples: avoidance of hard tasks, over-planning under-executing, energy crashes at same time, scope creep, task carry-over patterns.
 - Log as `PATTERN:` entry in Auto_Learnings.md (not `CORRECTION:`).

package/src/workflows/analyse-day.md

@@ -352,6 +352,11 @@ If tough day detected: "Consider running `/debug-resistance` to work through the
 
 ## 9. PROACTIVE SUGGESTIONS
 
+**CORRECTION Pre-Check:** Before routing patterns, scan Auto_Learnings.md for CORRECTION entries with Status: ACTIVE or DEFERRED. Surface these separately:
+- ACTIVE corrections (unresolved): "Correction '[name]' is still active (Conf [X]). Was this addressed today?"
+- DEFERRED corrections >30 days old: "Correction '[name]' was deferred on [date]. Still relevant?"
+- AI correction repeated 3x+ (Conf 5): "Recurring AI behavior issue: [name]. Propose inline fix or defer to `/review-week` 3H."
+
 Based on observations during analysis:
 1. [Pattern/suggestion] - Confidence: [X]
 2. [Pattern/suggestion] - Confidence: [X]
@@ -364,6 +369,7 @@ For any Auto_Learnings pattern with Confidence 3+, suggest the specific workflow
 - Identity drift (Conf 3+) → "Flag for `/review-week` identity audit."
 - Planning addiction (Conf 3+) → "You're over-planning again. Run `/debug-resistance` on the avoidance."
 - Scope creep (Conf 3+) → "Run `/refactor-plan` or `/overview` to reset scope."
+- AI correction repeated 3x+ (Conf 5) → "Structural AI issue. Propose inline fix via S11 Trigger A or defer to `/review-week` 3H."
 
 Don't just say "pattern detected." Route to the fix.
 

package/src/workflows/log.md

@@ -261,11 +261,31 @@ Output to user:
 ## 10. SELF-IMPROVEMENT
 
 If user corrected AI during this session:
+
+**Check first:** Did S11 Trigger A already handle this correction inline during the session? (Look for approved diff edits in conversation history.) If yes: log to Auto_Learnings only — skip the fix proposal.
+
+**If NOT already handled inline:**
 1. Acknowledge: "I made a mistake."
-2. Identify the rule violated.
-3. Append to Auto_Learnings.md:
-   `* [YYYY-MM-DD] CORRECTION: [Description]. Source: User correction. Confidence: 3.`
-4. If same correction seen 3x+, escalate confidence to 5.
+2. Identify the rule violated AND the source file.
+3. Propose fix using S11 diff format (see Global Rules S11 Trigger A).
+4. **STOP. Wait for user approval.**
+
+**In all cases, append structured CORRECTION to Auto_Learnings.md:**
+
+```markdown
+## CORRECTION: [Error Name]
+**Domain:** [AI Behavior / Workflow / Rule]
+**Error:** [What the AI did wrong]
+**Fix:** [What the correct behavior is]
+**Status:** [ACTIVE / RESOLVED / DEFERRED]
+**First Detected:** YYYY-MM-DD
+**Confidence:** 3
+**Last Updated:** YYYY-MM-DD
+### Evidence
+- YYYY-MM-DD: "[description]"
+```
+
+If same correction seen 3x+, escalate confidence to 5 and flag as structural.
 
 ---
 

package/src/workflows/review-week.md

@@ -221,6 +221,7 @@ Entries have different shelf lives based on their domain:
 | Health/Vitality | 180 days | "Energy crashes at 2pm", "Skips gym on Mondays" |
 | Preferences | 365 days | "Prefers voice-to-text", "Uses dark mode" |
 | Identity/Values | No expiry | "Driven by proving doubters wrong" |
+| AI Corrections | 90 days | "AI uses && on Windows", "AI overwrites Engine files" |
 
 If an entry exceeds its TTL without reconfirmation: reduce confidence by 1 and flag for review.
 If expired AND confidence ≤ 1: archive automatically.
@@ -231,6 +232,12 @@ If expired AND confidence ≤ 1: archive automatically.
 - Patterns confirmed 3+ times: promote confidence.
 - Chronic patterns (active 90+ days, confidence 4+): escalate — "This is structural, not situational."
 
+**4. CORRECTION Lifecycle:**
+- ACTIVE corrections absent 90+ days: reduce confidence by 1, set Status: MONITORING. "Fix appears to be working."
+- ACTIVE Conf 5 absent 90+ days: reduce to 4, flag: "Structural fix may have taken hold. Reducing severity."
+- DEFERRED corrections >30 days: surface explicitly — "Correction '[name]' deferred on [date]. Still relevant? Resolve or promote to ACTIVE."
+- RESOLVED corrections: no decay. Kept as permanent record.
+
 **Output:** Highlight top 3 active patterns for user awareness. Flag any entries that decayed or expired this week.
 
 ### 3G. Wins Extraction
@@ -492,11 +499,22 @@ Create `{VAULT_ROOT}/02_Areas/Engine/Quarterly Reviews/Quarter - [YYYY-QN].md` w
 ## 8. SELF-IMPROVEMENT
 
 If user corrected AI behavior during this workflow:
+
 1. Acknowledge: "I made a mistake."
-2. Identify which rule was violated.
-3. Append to `Auto_Learnings.md`:
-   `* [YYYY-MM-DD] CORRECTION: [Description]. Source: User correction during /review-week. Confidence: 3.`
-4. If same correction seen 3x+, escalate confidence to 5 and propose adding to Global Rules.
+2. Identify the violated rule AND the source file.
+3. Propose fix using S11 diff format:
+   ```
+   CORRECTION FIX:
+   - File: [path]
+   - OLD: "[current text]"
+   + NEW: "[proposed text]"
+   SAFETY CHECK: [Which Safe Modification gate this passes]
+   ```
+4. **STOP. Wait for user approval.** (Same gate as S11 Trigger A.)
+5. If approved: implement + log structured CORRECTION to Auto_Learnings.md (see S11 format).
+6. If deferred: log with `Status: DEFERRED`.
+7. Before proposing rule additions, pass the Rule Addition Gate (Global Rules S11).
+8. If same correction seen 3x+, escalate confidence to 5 and flag as structural.
 
 ---
 

package/src/workflows/update.md

@@ -58,7 +58,7 @@ cp "{VAULT_ROOT}/02_Areas/Engine/Metrics_Log.md" "$BACKUP_DIR/"
 cp "{VAULT_ROOT}/02_Areas/Engine/Someday_Maybe.md" "$BACKUP_DIR/"
 cp "{VAULT_ROOT}/02_Areas/Engine/Daily_Template.md" "$BACKUP_DIR/"
 ```
-Do NOT use `&&`. Run each `cp` separately.
+Do NOT use `&&`. Run each `cp` separately. If a file doesn't exist (e.g. V3 user), the `cp` will fail silently — that's expected.
 
 3. Create a git tag as a checkpoint:
 ```bash
@@ -99,6 +99,65 @@ If versions match: "You're already on the latest version. No update needed." →
 
 ---
 
+## 1.5 WHAT'S NEW
+
+Present this to the user after version detection confirms an update is available. Frame everything as **what you can do now**, not technical changelogs.
+
+**If upgrading from V3 to V4:**
+
+---
+
+**Your AI now challenges you (not just agrees).**
+The system has anti-RLHF behavioral rules. When you set a plan and drift from it, it will call it out. When you say "I'll do it tomorrow" without a specific first step, it pushes back. This is a feature, not a bug. Escalating Friction gives you 4 levels: from a gentle "Your plan says X, you're working on Y" to a hard block on destructive actions. You can always override levels 1-3 with a reason — the system creates friction, not walls. Every override gets logged. `/analyse-day` uses overrides to score whether your day was strategic or avoidant.
+
+**Your system learns from your corrections.**
+Say "don't do that" or "from now on do X" and the AI proposes a fix to its own workflow files in diff format. You approve or defer. Approved fixes take effect immediately. Corrections accumulate in Auto_Learnings.md with confidence scoring. Patterns get routed to specific fix workflows. The system literally rewrites itself based on how you use it. Weekly: `/review-week` Step 3H batches all unresolved corrections and proposes permanent fixes. Quarterly: rules get audited for dead weight and pruned.
+
+**Your daily flow is now a connected pipeline.**
+- `/morning` reads your identity, sets a Single Test for the day, checks energy, alerts you to active patterns.
+- `/plan-tomorrow` does dependency-aware task planning, zombie escalation (x3 = must break down or kill), and carry-over tracking from 2 days back.
+- `/log` captures session work with timestamps, syncs plan.md, detects commitments/decisions/people passively.
+- `/analyse-day` runs a brutal audit: Critical Path Check (was your work DIRECT, PREREQUISITE, or DRIFT?), Value Dysmorphia score, pattern detection.
+- Every workflow hands off to the next. The system tells you what to run next.
+
+**New workflows you didn't have:**
+- `/ignite` — Start a new project OR add a feature to an existing one (`--feature` flag). Deep interrogation (7 clusters, 45+ questions). Also: `--monetize` for full go-to-market pipeline.
+- `/debrief` — Run after any meeting or call. Extracts actions, decisions, commitments, and people automatically.
+- `/history` — See how your Identity, Strategy, and Goals have evolved over time (uses git log).
+- `/debug-resistance` — When you're avoiding something and you know it. Multi-step psychological diagnosis.
+- `/update` — This workflow. Safely migrates your Engine files when you update Mover OS.
+
+**Your Engine now tracks more.**
+- `Goals.md` — North Star targets at 90-day, 1-year, and 10-year horizons.
+- `Metrics_Log.md` — 30-day rolling performance data with baselines.
+- `Someday_Maybe.md` — Ideas that aren't current commitments but shouldn't be forgotten.
+- Entity memory (`03_Library/Entities/`) — People, Organizations, Places, Decisions. The system remembers who you've talked to and what was decided.
+- Input capture (`03_Library/Inputs/`) — Articles, Books, Videos you consume get logged with key takeaways.
+- `Active_Context.md` now has: Waiting For (what others owe you), Commitments (what you owe others), Backlog, My Customizations.
+
+**System infrastructure (runs silently).**
+- 6 hooks monitor your sessions: remind you to log, protect Engine files from accidental edits, warn on stale context, guard against `git add .`, back up transcripts before context compaction.
+- Hard-link architecture: edit workflows from any location, changes propagate everywhere.
+- 13 coding agents supported via `npx mover-os` installer.
+
+---
+
+**If upgrading within V4 (V4.x incremental):**
+
+| Version | What you can do now |
+|---|---|
+| V4.4 | Cross-workflow answer gating (can't skip questions by invoking a new workflow), PATTERN entries in Auto_Learnings, false causation guard |
+| V4.5 | Shared state bus (workflows detect if prerequisites ran), canonical daily note paths |
+| V4.6 | Decision persistence (decisions update files immediately), script routing (Library vs project), pattern-to-workflow routing |
+| V4.7 | Entity memory (People, Orgs, Decisions), commitments tracking, passive detection protocol, escalating friction |
+| V4.9+ | Critical path check in daily analysis, zombie task protection, monthly review tracking, comprehensive migration coverage, inline correction lifecycle |
+
+---
+
+**STOP. "Ready to sync workflows and migrate your Engine files? (Y/N)"**
+
+---
+
 ## 2. WORKFLOW & RULES UPDATE
 
 This is the safe part — workflows and rules don't contain user data.
@@ -108,11 +167,17 @@ This is the safe part — workflows and rules don't contain user data.
    - WAIT for user to confirm they've run it (or ask you to run it).
 
 2. If no installer available (manual install):
-   - List what changed between versions (read CHANGELOG from Global Rules)
+   - List what changed between versions (reference Section 1.5 above)
    - Guide user through manual file replacement
 
 3. Run `bash src/install/link.sh` to re-establish hard links.
 
+4. Check vault-level CLAUDE.md:
+   - Read `{VAULT_ROOT}/CLAUDE.md`. If it exists but is V3-era (no workflow table, no Engine file table, no "analyse-day"), replace it with the current bundle's CLAUDE.md.
+   - This is the project-level CLAUDE.md (instructions for this vault), NOT the global rules file (~/.claude/CLAUDE.md which is handled by link.sh).
+   - Show user: "Your vault's CLAUDE.md is outdated (V3). I'll replace it with the current version — this file describes how agents should interact with your vault, it doesn't contain personal data."
+   - WAIT FOR CONFIRMATION before replacing.
+
 Output: "[X] workflows synced. [Y] skills updated. Rules updated to V{new}."
 
 ---
@@ -169,6 +234,9 @@ Read the file. Check for:
 |---------|----------|-------|
 | Version header | V4.6 | Contains `**Version:**` ? |
 | Title is "MOVER DOSSIER" not "ARCHITECT DOSSIER" | V4.6 | Check H1 heading |
+| Specific Knowledge section (S3.5) | V4.0 | Contains "Specific Knowledge" or "Naval" heading? |
+| Survival Capital section (S6) | V4.0 | Contains "SURVIVAL CAPITAL" heading? |
+| Constraints section (S7) | V4.0 | Contains "CONSTRAINTS" heading? |
 
 ### 3E. Goals.md
 
@@ -180,12 +248,58 @@ Read the file. Check for:
 
 | Section | Added In | Check |
 |---------|----------|-------|
-| PATTERN entry type documented | V4.4 | Contains "PATTERN:" in any entry or header? |
+| Two entry types documented (header) | V4.4 | Contains "CORRECTION" AND "PATTERN" in headers or entry examples? |
+| PATTERN entry structure | V4.4 | Contains "Type:", "Trigger:", "Counter-Action:" fields? |
 
 ### 3G. Metrics_Log.md
 
 Check if file exists. If not, it was added in V4.0.
 
+### 3H. Someday_Maybe.md
+
+Check if file exists at `02_Areas/Engine/Someday_Maybe.md`. If not, it was added in V4.0.
+
+### 3I. Daily_Template.md
+
+Read the file. Check for:
+
+| Section | Added In | Check |
+|---------|----------|-------|
+| Food tracking | V4.7 | Contains "Food" in the Log section? |
+| Sleep tracking | V4.7 | Contains "Sleep:" or "Sleep Log" in the Log section? |
+| Rudder section | V4.4 | Contains "Rudder" heading? |
+| Key Insights section | V4.4 | Contains "Key Insights" heading? |
+| Metrics section | V4.0 | Contains "Metrics" heading? |
+
+### 3J. Library Structure
+
+Check if these directories and files exist:
+
+| Path | Added In | Check |
+|------|----------|-------|
+| `03_Library/Entities/` | V4.0 | Directory exists? |
+| `03_Library/Entities/People/` | V4.0 | Directory exists? |
+| `03_Library/Entities/Organizations/` | V4.0 | Directory exists? |
+| `03_Library/Entities/Places/` | V4.0 | Directory exists? |
+| `03_Library/Entities/Decisions/` | V4.0 | Directory exists? |
+| `03_Library/Entities/Entities MOC.md` | V4.0 | File exists? |
+| `03_Library/Inputs/` | V4.0 | Directory exists? |
+| `03_Library/Inputs/Articles/` | V4.0 | Directory exists? |
+| `03_Library/Inputs/Books/` | V4.0 | Directory exists? |
+| `03_Library/Inputs/Videos/` | V4.0 | Directory exists? |
+| `03_Library/Inputs/Inputs MOC.md` | V4.0 | File exists? |
+
+### 3K. Vault CLAUDE.md
+
+Check if `{VAULT_ROOT}/CLAUDE.md` exists and contains V4 content:
+
+| Check | Added In | How |
+|-------|----------|-----|
+| V4 workflow table | V4.0 | Contains "analyse-day" in a workflow table? |
+| Engine file table | V4.0 | Contains "Active_Context.md" in an Engine table? |
+
+If the file exists but looks like V3 (no workflow table, no Engine table): flag for refresh.
+
 ---
 
 ## 4. MIGRATION REPORT
@@ -206,8 +320,29 @@ Identity_Prime.md:
 Active_Context.md:
   ✓ Strategic Verdict
   ✗ MISSING: Workflow State (V4.5)
+  ✗ MISSING: Identity Snapshot (V4.7)
+  ✗ MISSING: Backlog (V4.7)
+
+Mover_Dossier.md:
+  ✓ Skills Inventory
+  ✗ MISSING: Specific Knowledge (V4.0)
+
+New files needed:
+  ✗ Goals.md (V4.0)
+  ✗ Metrics_Log.md (V4.0)
+  ✗ Someday_Maybe.md (V4.0)
 
-[...etc for all files]
+Template updates:
+  ✗ Daily_Template.md: missing Food/Sleep, Rudder, Key Insights
+
+Library structure:
+  ✗ 03_Library/Entities/ (directories + templates)
+  ✗ 03_Library/Inputs/ (directories + templates)
+
+Auto_Learnings.md:
+  ✗ MISSING: PATTERN entry type
+
+[...check all sections from 3A-3K]
 
 Migrations needed: [N]
 ```
@@ -423,6 +558,254 @@ Read the full Identity_Prime.md. Derive the profile from existing content:
 "Adding a hidden metadata tag to Identity_Prime.md for other workflows to reference. Derived from your existing profile: [show values]. Accurate?"
 WAIT FOR CONFIRMATION.
 
+### 5Q. Metrics_Log.md (if file missing entirely)
+
+No interview needed — structural file populated by /analyse-day. Create from template:
+
+```markdown
+# Metrics Log
+
+**Purpose:** 30-day rolling performance data. Archived monthly by `/review-week`.
+**Written By:** `/analyse-day`, `/review-week`
+**Read By:** `/plan-tomorrow`, `/review-week`, `/analyse-day`
+
+---
+
+## Current Period
+
+### Weekly Summary
+
+| Week | Completion % | Deep Work Avg | Energy Avg | Zombie Tasks | Single Test Hit Rate |
+| :--- | :--- | :--- | :--- | :--- | :--- |
+| | | | | | |
+
+### Daily Entries
+
+*Appended by `/analyse-day`. One row per day.*
+
+| Date | Planned | Done | Rate | Deep Work | Energy | Single Test | Notes |
+| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
+| | | | | | | | |
+
+---
+
+## Baseline
+
+*Established by `/review-week` Monthly Mode after Month 1 (20+ daily entries).*
+
+| Metric | Value | Month Established |
+| :--- | :--- | :--- |
+| Completion Rate | — | — |
+| Deep Work Avg (h) | — | — |
+| Energy Avg (1-10) | — | — |
+| Single Test Hit Rate | — | — |
+| Zombie Task Avg | — | — |
+
+---
+
+## Archive
+
+*Moved here by `/review-week` at month end. Never deleted.*
+```
+
+"Creating Metrics_Log.md — your `/analyse-day` workflow will populate this with daily performance data."
+WAIT FOR CONFIRMATION.
+
+### 5R. Someday_Maybe.md (if file missing entirely)
+
+No interview needed — structural file. Create from template:
+
+```markdown
+# Someday / Maybe
+
+**Last Updated:** [today]
+
+**Purpose:** Ideas, projects, and intentions that are NOT current commitments. Not now, but not never.
+**Written By:** User directly, `/log` (detected from session), `/ignite` (if user declines to start now)
+**Read By:** `/review-week` monthly mode (triage — kill, activate, or keep)
+**Rule:** Never let this list pressure you. These are options, not obligations.
+
+---
+
+## Projects
+
+*Ideas that could become projects. Not started, not committed.*
+
+- [ ] [Project idea] — [Why it's interesting] — [What would trigger activation] — [Added: YYYY-MM-DD]
+
+---
+
+## Experiments
+
+*Things to try once. Single actions, not projects.*
+
+- [ ] [Experiment] — [Hypothesis] — [Added: YYYY-MM-DD]
+
+---
+
+## Learn
+
+*Skills, topics, or domains to explore when there's space.*
+
+- [ ] [Topic/skill] — [Why] — [Added: YYYY-MM-DD]
+
+---
+
+## Buy / Build
+
+*Tools, gear, or systems worth getting eventually.*
+
+- [ ] [Item] — [Why] — [Approx cost or effort] — [Added: YYYY-MM-DD]
+
+---
+
+## Review Protocol
+
+During `/review-week` Monthly Mode, triage this list:
+- **Activate:** Move to `01_Projects/` and run `/ignite`
+- **Kill:** Remove if interest is gone — no guilt
+- **Keep:** Leave it. If it's been here 6+ months, consider killing it
+
+Items older than 6 months without a pulse get flagged: "This has been Someday for [N] months. Kill it or give it a date?"
+```
+
+"Creating Someday_Maybe.md — a parking lot for future ideas. `/log` captures these automatically, `/review-week` triages them monthly."
+WAIT FOR CONFIRMATION.
+
+### 5S. Auto_Learnings.md PATTERN Structure (if only CORRECTION entries exist)
+
+No interview needed. The file exists but may only document CORRECTION entries (V3 format). Append a PATTERN example after the existing entries:
+
+1. Read the current file. If it already has PATTERN entries with Type/Trigger/Counter-Action fields, skip.
+2. If not, ensure the file header mentions both types: "Two entry types: CORRECTION (AI behavior fixes) and PATTERN (user behavioral patterns)."
+3. Append a PATTERN template entry:
+
+```markdown
+## PATTERN: [Pattern Name]
+**Type:** [Avoidance / Energy / Scope / Execution / Habit]
+**Domain:** [Business/Vitality/Mind]
+**Hypothesis:** [Vx]
+**Suspected Variable:** [Variable Name]
+**Trigger:** [What causes this pattern to fire]
+**Counter-Action:** [What to do instead]
+**Status:** [ACTIVE/MONITORING/RESOLVED]
+**First Detected:** [Date]
+**Confidence:** 1
+**Severity:** LOW
+**Last Updated:** [Date]
+
+### Evidence
+- [Date]: "Placeholder — `/analyse-day` will populate with real observations."
+```
+
+"Updating Auto_Learnings.md with PATTERN entry format. Your `/analyse-day` workflow now detects behavioral patterns (avoidance, energy crashes, scope creep) in addition to AI corrections."
+WAIT FOR CONFIRMATION.
+
+### 5T. Daily_Template.md Updates (if missing sections)
+
+No interview needed — template updates. Read the current Daily_Template.md and add any missing sections:
+
+1. **Food + Sleep tracking** (if missing): Add after the freeform log line (`- 09:00 -`):
+```markdown
+**Food:**
+
+**Sleep:** Total: | Deep:
+```
+
+2. **Rudder section** (if missing): Add after Session Log:
+```markdown
+## Rudder
+
+*AI-generated analysis. Written by `/analyse-day`.*
+
+- **Verdict:** [Pass/Fail — with citations]
+- **Volume Check:** [Did we hit daily volume?]
+- **Battery Status:** [Energy assessment]
+- **System Tweak:** [Fix for tomorrow]
+```
+
+3. **Key Insights section** (if missing): Add after Rudder:
+```markdown
+## Key Insights
+
+*AI-generated tactical insights for tomorrow. Written by `/analyse-day`.*
+```
+
+4. **Metrics section** (if missing): Add after Key Insights:
+```markdown
+## Metrics
+
+| Metric | Value | Notes |
+| :--- | :--- | :--- |
+| Tasks Planned | [0] | From Focus + Tasks |
+| Tasks Completed | [0] | Checked off |
+| Completion Rate | [0%] | Auto-calculated |
+| Zombie Tasks | [0] | Carried 2+ days |
+| Deep Work Actual | [0h] | From Schedule |
+| Single Test Hit? | [Y/N] | Did you win the day? |
+| Energy Avg | [0] | Self-reported 1-10 |
+```
+
+"Updating your Daily_Template.md with new sections. Future daily notes will include: food/sleep tracking, metrics dashboard, rudder analysis, and key insights."
+WAIT FOR CONFIRMATION. Only write sections that are actually missing.
+
+### 5U. Library Structure (if directories/templates missing)
+
+No interview needed — structural scaffolding. Create any missing directories and template files:
+
+**Entities system** (if `03_Library/Entities/` missing):
+1. Create directories: `Entities/People/`, `Entities/Organizations/`, `Entities/Places/`, `Entities/Decisions/`
+2. Copy templates from bundle: `_TEMPLATE.md` into each subdirectory
+3. Copy `Entities MOC.md` from bundle
+
+**Inputs system** (if `03_Library/Inputs/` missing):
+1. Create directories: `Inputs/Articles/`, `Inputs/Books/`, `Inputs/Videos/`
+2. Copy templates from bundle: `_TEMPLATE.md` into each subdirectory
+3. Copy `Inputs MOC.md` from bundle
+
+"Adding Library structure for entity memory and input capture. These enable `/log` to track people, organizations, articles, and decisions automatically."
+WAIT FOR CONFIRMATION.
+
+### 5V. Dossier Specific Knowledge (if S3.5 missing)
+
+Interview needed — this requires understanding the user's unique strengths.
+
+"V4 added a 'Specific Knowledge' section to your Dossier — based on Naval Ravikant's framework. Two questions:"
+
+1. "What do you do that feels like play to you but looks like work to everyone else?"
+2. "What's the intersection of that with your current industry or business?"
+
+**CHECKPOINT:** "Here's what I'll add as Section 3.5 in Mover_Dossier.md: [draft section]. Accurate?"
+WAIT FOR CONFIRMATION.
+
+### 5W. Dossier Missing Sections (Survival Capital, Constraints)
+
+If `SURVIVAL CAPITAL` section missing:
+
+No interview needed — user can fill in later. Append scaffold:
+
+```markdown
+## 6. SURVIVAL CAPITAL
+
+*   **Runway:** [X] months at $[Monthly Burn]
+*   **Emergency Fund:** $[Amount]
+*   **Fallback Plan:** [What you'd do if this fails]
+```
+
+If `CONSTRAINTS` section missing:
+
+```markdown
+## 7. CONSTRAINTS (The Box)
+
+*   **Time:** [Hours/week available]
+*   **Energy Windows:** Peak: [Times], Dead: [Times]
+*   **Location:** [Remote/On-site, Travel limitations]
+*   **Non-Negotiables:** [Sabbath, Family time, Health appointments]
+```
+
+"Adding Survival Capital and Constraints sections to your Dossier. Fill these in when you have a moment — they help the system give realistic recommendations."
+WAIT FOR CONFIRMATION.
+
 ---
 
 ## 6. APPLY MIGRATIONS
@@ -445,8 +828,15 @@ git add 02_Areas/Engine/Active_Context.md
 git add 02_Areas/Engine/Strategy.md
 git add 02_Areas/Engine/Mover_Dossier.md
 git add 02_Areas/Engine/Goals.md
+git add 02_Areas/Engine/Auto_Learnings.md
+git add 02_Areas/Engine/Metrics_Log.md
+git add 02_Areas/Engine/Someday_Maybe.md
+git add 02_Areas/Engine/Daily_Template.md
+git add 03_Library/Entities/
+git add 03_Library/Inputs/
+git add CLAUDE.md
 ```
-Only add files that were actually modified.
+Only add files that were actually modified or created.
 
 ```bash
 git commit -m "update: migrate Engine files to V{new}"
@@ -502,6 +892,11 @@ Based on what was migrated:
 1. If Vitality/Spiritual sections were just added: "Your Identity now covers three domains. Run `/morning` to visceralize the full picture." - Confidence: 4
 2. If Workflow State was added: "Cross-workflow prerequisites are now active. `/plan-tomorrow` will check for `/review-week` on Sundays." - Confidence: 5
 3. If Goals.md was just created: "Review your Goals with fresh eyes tomorrow. First-draft goals from a migration interview are usually 70% right." - Confidence: 3
+4. If Daily_Template was updated: "Your daily notes template now includes food/sleep tracking and metrics. Next daily note you create will have the new format." - Confidence: 5
+5. If Entities/Inputs structure was created: "Entity memory is active. When you mention people or organizations, the system will offer to create entity notes." - Confidence: 4
+6. If Someday_Maybe was created: "Someday/Maybe list is now available. Ideas that aren't commitments go here instead of cluttering your active plans." - Confidence: 4
+7. If Metrics_Log was created: "Run `/analyse-day` tonight to start populating your Metrics Log. After 20+ entries, the system establishes performance baselines." - Confidence: 4
+8. If Auto_Learnings PATTERN type was added: "Your system now tracks behavioral patterns (avoidance, energy, scope creep) — not just AI corrections. `/analyse-day` feeds this automatically." - Confidence: 5
 
 "Want me to address any of these?"