moveros 4.0.8 → 4.1.0

package/src/system/V4_CONTEXT.md

@@ -1,262 +1,262 @@
-# V4 CONTEXT: Workflow Engineering Implementation State
-
-**Purpose:** Persistent memory for any AI instance working on Mover OS V4.
-**Last Updated:** 2026-02-23
-**Paradigm:** Workflow Engineering (NOT Prompt Engineering)
-
----
-
-## 0. THE PARADIGM SHIFT (READ FIRST)
-
-| Old (2024) | New (2025/2026) |
-|------------|-----------------|
-| **Prompt Optimization:** Perfect instruction writing | **System Optimization:** Route, verify, ground |
-| **Single Shot / CoT:** Step-by-step thinking | **Inference-Time Search:** Multiple paths before output |
-| **Vibe Check:** Manual output review | **Automated Benchmarking:** Golden Template comparison |
-| **Linear:** Step 1 → 2 → Output | **Agentic:** Router → Experts → Verifier → Output |
-
-> **We don't write prompts. We build a Testing Harness that evolves prompts.**
-
----
-
-## 1. RESEARCH SOURCES
-
-| File | Lines | Purpose |
-|------|-------|---------|
-| `V4_Research_Synthesis.md` | 899 | Foundation research (2025 academics) |
-| `groundbreaking_2025.md` | 751 | 10 Critical Findings (paradigm shift) |
-| `INTERROGATION_V4.md` | 4000+ | 410+ design decisions |
-| `plan.md` (Phase 11) | 200+ | Implementation roadmap |
-| `V4_ARCHITECTURE_DEEP_DESIGN.md` | 1013 | Complete V4 architecture |
-| `WORKFLOW_TEMPLATE.md` | 309 | V4 master workflow template |
-| `WORKFLOW_PRINCIPLES.md` | 489 | 11 governing principles |
-| `WORKFLOW_MASTER_PROTOCOL.md` | 168 | Supreme authority protocol |
-
----
-
-## 2. CRITICAL FINDINGS (10 Total)
-
-| # | Finding | Source | ROI |
-|---|---------|--------|-----|
-| 1 | Inference-Time Scaling > Longer Prompts | OpenAI o3, DeepSeek-R1 | Request reasoning density |
-| 2 | MoE Routing for Workflow Selection | GPT-5 MoE, LLMoE | Router → Expert workflows |
-| 3 | Coactive Learning (in-session) | EMNLP 2025 | Learn from user edits |
-| 4 | Structured Data = 40-60% fewer hallucinations | Lakera AI | JSON schemas required |
-| 5 | Temporal Knowledge Graphs | Zep (+18.5%) | Daily Notes = memory kernel |
-| 6 | Prompt Injection Defense | ACL 2025 | Context isolation |
-| 7 | Selective Reasoning Activation | Raschka | Tier by cost |
-| 8 | Ensemble Verification (Jekyll/Hyde) | +9.98% | Risk-stratified |
-| 9 | Confidence Calibration | ACL 2025 | Historical calibration |
-| 10 | Dynamic Prompt Adaptation | 76pp variance | Position/Length/Rep |
-
----
-
-## 3. KEY ARCHITECTURAL DECISIONS (FINAL)
-
-| Decision | Reason |
-|----------|--------|
-| **REMOVE Mercenary Mode** | Confuses LLM, premature execution |
-| **MERGE `/checkup` → `/log`** | Consolidate session logging |
-| **MERGE `/context` → `/analyse-day`** | Redundant |
-| **MERGE `/project-notes` → `/overview`** | Consolidate project context |
-| **ADD `/overview`** | Project resume (new workflow) |
-| **Single Architect persona** | Adaptive to context, no mode switching |
-| **Confidence scoring 1-5** | Threshold: < 3 = refuse to assert |
-| **Source citation [file:line]** | Grounding and verification anchor |
-| **Verification loops** | Risk-tiered: HIGH=user confirm, MEDIUM=CISC, LOW=direct |
-| **//turbo mode** | Skip non-HIGH confirmations for session |
-| **12k char limit** | Use skill sub-modules if needed |
-| **Memory Hierarchy P0-P5** | Auto_Learnings > Active_Context > Strategy > Identity > project_state > Dailies |
-| **Areeb's debugging framework** | Self-Critique > Verbose Logging > 5-Why > Differential Diagnosis |
-| **Three domains in Identity** | Work, Vitality, Faith/Purpose |
-| **Proactive surfacing** | Suggest, never auto-execute |
-| **Answer gating** | No handoff until all questions answered |
-| **Library dedup protocol** | Check existing files before creating new ones |
-
----
-
-## 4. IMPLEMENTATION PROGRESS
-
-### Phase 11.0: Foundation (COMPLETE)
-- [x] WORKFLOW_TEMPLATE.md (V4 master template)
-- [x] WORKFLOW_PRINCIPLES.md (11 principles)
-- [x] WORKFLOW_MASTER_PROTOCOL.md (supreme authority)
-- [x] V4_ARCHITECTURE_DEEP_DESIGN.md (complete architecture)
-
-### Phase 11.1: Research (COMPLETE)
-- [x] V4_Research_Synthesis.md
-- [x] groundbreaking_2025.md
-- [x] Paradigm shift identified
-- [x] Claude Code native capabilities researched (hooks, sub-agents, CLAUDE.md)
-
-### Phase 11.2: Global Rules V4 (COMPLETE)
-- [x] Mover_Global_Rules.md rewritten to V4
-- [x] Mercenary Mode removed
-- [x] Grounding & Citation (S2) added
-- [x] Verification & Confidence (S3) added
-- [x] Memory Hierarchy (S0.5) added
-- [x] Proactive Surfacing (S12) added
-- [x] Answer Gating (S13) added
-- [x] Library Protocol (S14) added
-- [x] AI Behavior (S17) added
-- [x] Areeb's 4-step debugging framework integrated into Coding (S4)
-- [x] //turbo mode added
-
-### Phase 11.3: Critical Workflow Rewrites (COMPLETE)
-- [x] `/log` V4 (absorbs /checkup, verbose session logs, CISC)
-- [x] `/analyse-day` V4 (absorbs /context, variable interrogation, scientific patterns)
-- [x] `/plan-tomorrow` V4 (zombie audit, dependency awareness, project scope)
-- [x] `/review-week` V4 (system improvement point, monthly mode, sample size)
-- [x] `/overview` V4 (NEW - absorbs /project-notes, staleness detection, portfolio mode)
-
-### Phase 11.4: Remaining Workflow Rewrites (COMPLETE)
-- [x] `/primer` V4 (session primer, pattern check, energy-based)
-- [x] `/harvest` V4 (mandatory Library dedup, MOC genesis, upward linking)
-- [x] `/context` V4 (DEPRECATED - merged into /analyse-day)
-- [x] `/ignite` V4 (single persona, Library pre-search, leverage audit)
-- [x] `/setup` V4 (Bank G Vitality, Bank H Faith/Purpose, Goals.md, Perpetual Protocol)
-- [x] `/walkthrough` V4 (14 workflows, V4 features, three domains)
-- [x] `/init-plan` V4 (dynamic logic, stack-aware, CISC)
-- [x] `/debug-resistance` V4 (multi-step interrogation, Auto_Learnings integration)
-- [x] `/pivot-strategy` V4 (sample size validation, HIGH-RISK gate, version tracking)
-- [x] `/reboot` V4 (staleness tiers, recurring reboot detection, escalation)
-- [x] `/refactor-plan` V4 (system-triggerable, surgical edit, CISC)
-- [x] `/project-notes` V4 (DEPRECATED - merged into /overview)
-
-### Phase 11.4.1: V4 Compliance Fixes (COMPLETE)
-- [x] Tested all 15 workflows against V4 checklist (13 items) + 11 principles
-- [x] Fixed all 15 workflows: added self-improvement, proactive suggestions, answer gating, CISC, bottom bookend
-- [x] Fixed broken `[V4_ARCHITECTURE]` references (5 occurrences across 4 workflows)
-- [x] Fixed memory hierarchy gaps (reboot P0 order, overview P3/P5, init-plan P0, refactor-plan P0)
-- [x] Fixed Dossier reference in debug-resistance.md
-- [x] Added workflow registry (15 active + 2 deprecated) to Global Rules
-- [x] Added supporting flow paths, three domains, paradigm statement to Global Rules
-- [x] Daily Note Template V4: added `## Session Log` + `## Metrics` sections
-- [x] Created `Goals.md` template in Engine
-- [x] Created `Metrics_Log.md` template in Engine
-- [x] Removed Co-Authored-By attributions from all V4 commits
-- [x] Disabled future attribution in Claude Code settings
-
-### Phase 11.5: Golden Templates (PENDING)
-- [ ] `Golden_Templates/` folder
-- [ ] `Golden_AnalyseDay.json`
-- [ ] `Golden_PlanTomorrow.json`
-- [ ] `Golden_ReviewWeek.json`
-
-### Phase 11.6: Benchmarking System (PENDING)
-- [ ] `V4_Benchmark_Protocol.md`
-- [ ] Sample test data
-- [ ] LLM-as-Judge prompt
-
-### Phase 11.7a: Claude Code Hooks (COMPLETE)
-- [x] Stop hook: session-log-reminder (reminds /log after 5+ exchanges)
-- [x] PreToolUse hook: engine-protection (confirms before writing core Engine files)
-- [x] `.claude/settings.json` with hooks config
-- [x] `.claude/hooks/` directory with executable scripts
-- [x] No jq dependency (pure grep/sed for Windows compatibility)
-
-### Phase 11.7b: Advanced Features (PARTIAL)
-- [x] `/history` workflow (git-based Engine file tracking) — COMPLETE
-- [ ] MCP integration (Health Connect, Strong App, Context7, Chrome)
-- [ ] Skills architecture for large workflows (/ignite, /setup)
-- [ ] Session quality auto-evaluation hook
-
----
-
-## 5. V4 WORKFLOW PATTERNS (All workflows now follow these)
-
-1. **Pre-flight:** Memory hierarchy load (P0-P5), rules confirmation
-2. **Prerequisite check:** Verify required files exist, chain prerequisites
-3. **Context injection:** Grounded data extraction from Engine files
-4. **Execution:** All claims cite [file:line] or [INFERRED: reason]
-5. **CISC self-verification:** 90%+ accuracy check before output
-6. **Risk-tiered verification:** HIGH=user confirm, MEDIUM=CISC, LOW=direct
-7. **Self-improvement:** Log corrections to Auto_Learnings.md
-8. **Proactive surfacing:** Pattern-based suggestions with confidence scores
-9. **Answer gating:** No handoff until all questions answered
-10. **Handoff (bottom bookend):** Sources, confidence range, next workflow
-
----
-
-## 6. RULES FOR FUTURE AI INSTANCES
-
-1. **Read this file first** before any workflow changes
-2. **Paradigm:** We do Workflow Engineering, not Prompt Engineering
-3. **Never recreate Mercenary Mode** - deliberately removed
-4. **Confidence scores:** 1-5 scale, threshold 3
-5. **Citations:** [file:line] or [INFERRED: reason] required
-6. **Update progress** in Section 4 after completing tasks
-7. **Append-only:** Never delete from plan.md, project_state.md, or Auto_Learnings.md
-8. **Three domains:** Identity covers Work, Vitality, Faith/Purpose
-9. **Areeb framework:** Self-Critique > Verbose Logging > 5-Why > Differential Diagnosis
-10. **Windows:** Never use && in commands
-
----
-
-## 7. FILES CREATED/MODIFIED IN V4
-
-| File | Path | Status |
-|------|------|--------|
-| Mover_Global_Rules.md | `src/system/` | Rewritten to V4 |
-| V4_CONTEXT.md | `src/system/` | Updated |
-| log.md | `src/workflows/` | Rewritten to V4 |
-| analyse-day.md | `src/workflows/` | Rewritten to V4 |
-| plan-tomorrow.md | `src/workflows/` | Rewritten to V4 |
-| review-week.md | `src/workflows/` | Rewritten to V4 |
-| overview.md | `src/workflows/` | NEW (V4) |
-| morning.md | `src/workflows/` | Rewritten to V4 |
-| harvest.md | `src/workflows/` | Rewritten to V4 |
-| context.md | `src/workflows/` | DEPRECATED |
-| ignite.md | `src/workflows/` | Rewritten to V4 |
-| setup.md | `src/workflows/` | Rewritten to V4 |
-| walkthrough.md | `src/workflows/` | Rewritten to V4 |
-| init-plan.md | `src/workflows/` | Rewritten to V4 |
-| debug-resistance.md | `src/workflows/` | Rewritten to V4 |
-| pivot-strategy.md | `src/workflows/` | Rewritten to V4 |
-| reboot.md | `src/workflows/` | Rewritten to V4 |
-| refactor-plan.md | `src/workflows/` | Rewritten to V4 |
-| project-notes.md | `src/workflows/` | DEPRECATED |
-
----
-
-## 8. V4.x INCREMENTAL UPDATES (Post-Launch)
-
-Changes since initial V4 release (2026-02-19):
-
-| Version | Date | Changes |
-|---------|------|---------|
-| 4.0.9 | 2026-02-21 | All 17 workflows intelligence-fixed, workflow compliance audit |
-| 4.0.10 | 2026-02-22 | 13-agent installer (Codex split, Gemini CLI), skill category multi-select, tiered installs |
-| 4.0.11 | 2026-02-23 | Global Rules trimmed (agent-agnostic skills, removed VoltAgent/CHANGELOG/Areeb refs) |
-| 4.0.12 | 2026-02-23 | Hook double-fire fix, installer deep-merge for hooks, tiered workflow reads, Identity Snapshot + Active Patterns in Active_Context |
-| 4.0.13 | 2026-02-23 | Install.js TDZ crash fix, S0 Active_Context read mandate, /update detection + migration handlers |
-| 4.0.14 | 2026-02-23 | /update V3→V4 migration audit (4 detection sections, 7 migration handlers) |
-| 4.0.15 | 2026-02-23 | Correction lifecycle (S11 rewrite, Safe Modification Protocol), /update What's New section, project_state cleanup |
-| 4.0.16 | 2026-02-23 | Tier reclassification (Recommended/Supported/Compatible/Basic), interactive walkthrough rewrite, full documentation update |
-
-### Key Changes Since Initial V4
-
-- **13 agents** (was 11): Codex split into CLI + Cloud, Gemini CLI added
-- **`/primer` renamed to `/morning`** — works the same, better name
-- **`/history` workflow** complete (was PENDING)
-- **`/debrief` workflow** added (post-meeting capture)
-- **38 skills** (was 40+ — consolidated during audit)
-- **Install command:** `npx moveros` (was `npx mover-os`)
-- **Tier labels removed** — table shows Agent + What Gets Installed (no tier column). Claude Code callout for deepest integration.
-- **Correction lifecycle:** S11 Trigger A provides same-session inline fix path with diff format
-- **Safe Modification Protocol:** 7 guardrails for self-editing
-- **Interactive walkthrough:** STOP gates, live demos, capability groups
-- **Missing workflows in file table:** /debrief, /history, /update (added below)
-
-### Additional Files (not in original Section 7)
-
-| File | Path | Status |
-|------|------|--------|
-| debrief.md | `src/workflows/` | NEW (V4) |
-| history.md | `src/workflows/` | NEW (V4) |
-| update.md | `src/workflows/` | NEW (V4) |
-
----
-
-**END OF V4 CONTEXT FILE**
-*Update after each implementation session.*
+# V4 CONTEXT: Workflow Engineering Implementation State
+
+**Purpose:** Persistent memory for any AI instance working on Mover OS V4.
+**Last Updated:** 2026-02-23
+**Paradigm:** Workflow Engineering (NOT Prompt Engineering)
+
+---
+
+## 0. THE PARADIGM SHIFT (READ FIRST)
+
+| Old (2024) | New (2025/2026) |
+|------------|-----------------|
+| **Prompt Optimization:** Perfect instruction writing | **System Optimization:** Route, verify, ground |
+| **Single Shot / CoT:** Step-by-step thinking | **Inference-Time Search:** Multiple paths before output |
+| **Vibe Check:** Manual output review | **Automated Benchmarking:** Golden Template comparison |
+| **Linear:** Step 1 → 2 → Output | **Agentic:** Router → Experts → Verifier → Output |
+
+> **We don't write prompts. We build a Testing Harness that evolves prompts.**
+
+---
+
+## 1. RESEARCH SOURCES
+
+| File | Lines | Purpose |
+|------|-------|---------|
+| `V4_Research_Synthesis.md` | 899 | Foundation research (2025 academics) |
+| `groundbreaking_2025.md` | 751 | 10 Critical Findings (paradigm shift) |
+| `INTERROGATION_V4.md` | 4000+ | 410+ design decisions |
+| `plan.md` (Phase 11) | 200+ | Implementation roadmap |
+| `V4_ARCHITECTURE_DEEP_DESIGN.md` | 1013 | Complete V4 architecture |
+| `WORKFLOW_TEMPLATE.md` | 309 | V4 master workflow template |
+| `WORKFLOW_PRINCIPLES.md` | 489 | 11 governing principles |
+| `WORKFLOW_MASTER_PROTOCOL.md` | 168 | Supreme authority protocol |
+
+---
+
+## 2. CRITICAL FINDINGS (10 Total)
+
+| # | Finding | Source | ROI |
+|---|---------|--------|-----|
+| 1 | Inference-Time Scaling > Longer Prompts | OpenAI o3, DeepSeek-R1 | Request reasoning density |
+| 2 | MoE Routing for Workflow Selection | GPT-5 MoE, LLMoE | Router → Expert workflows |
+| 3 | Coactive Learning (in-session) | EMNLP 2025 | Learn from user edits |
+| 4 | Structured Data = 40-60% fewer hallucinations | Lakera AI | JSON schemas required |
+| 5 | Temporal Knowledge Graphs | Zep (+18.5%) | Daily Notes = memory kernel |
+| 6 | Prompt Injection Defense | ACL 2025 | Context isolation |
+| 7 | Selective Reasoning Activation | Raschka | Tier by cost |
+| 8 | Ensemble Verification (Jekyll/Hyde) | +9.98% | Risk-stratified |
+| 9 | Confidence Calibration | ACL 2025 | Historical calibration |
+| 10 | Dynamic Prompt Adaptation | 76pp variance | Position/Length/Rep |
+
+---
+
+## 3. KEY ARCHITECTURAL DECISIONS (FINAL)
+
+| Decision | Reason |
+|----------|--------|
+| **REMOVE Mercenary Mode** | Confuses LLM, premature execution |
+| **MERGE `/checkup` → `/log`** | Consolidate session logging |
+| **MERGE `/context` → `/analyse-day`** | Redundant |
+| **MERGE `/project-notes` → `/overview`** | Consolidate project context |
+| **ADD `/overview`** | Project resume (new workflow) |
+| **Single Architect persona** | Adaptive to context, no mode switching |
+| **Confidence scoring 1-5** | Threshold: < 3 = refuse to assert |
+| **Source citation [file:line]** | Grounding and verification anchor |
+| **Verification loops** | Risk-tiered: HIGH=user confirm, MEDIUM=CISC, LOW=direct |
+| **//turbo mode** | Skip non-HIGH confirmations for session |
+| **12k char limit** | Use skill sub-modules if needed |
+| **Memory Hierarchy P0-P5** | Auto_Learnings > Active_Context > Strategy > Identity > project_state > Dailies |
+| **Areeb's debugging framework** | Self-Critique > Verbose Logging > 5-Why > Differential Diagnosis |
+| **Three domains in Identity** | Work, Vitality, Faith/Purpose |
+| **Proactive surfacing** | Suggest, never auto-execute |
+| **Answer gating** | No handoff until all questions answered |
+| **Library dedup protocol** | Check existing files before creating new ones |
+
+---
+
+## 4. IMPLEMENTATION PROGRESS
+
+### Phase 11.0: Foundation (COMPLETE)
+- [x] WORKFLOW_TEMPLATE.md (V4 master template)
+- [x] WORKFLOW_PRINCIPLES.md (11 principles)
+- [x] WORKFLOW_MASTER_PROTOCOL.md (supreme authority)
+- [x] V4_ARCHITECTURE_DEEP_DESIGN.md (complete architecture)
+
+### Phase 11.1: Research (COMPLETE)
+- [x] V4_Research_Synthesis.md
+- [x] groundbreaking_2025.md
+- [x] Paradigm shift identified
+- [x] Claude Code native capabilities researched (hooks, sub-agents, CLAUDE.md)
+
+### Phase 11.2: Global Rules V4 (COMPLETE)
+- [x] Mover_Global_Rules.md rewritten to V4
+- [x] Mercenary Mode removed
+- [x] Grounding & Citation (S2) added
+- [x] Verification & Confidence (S3) added
+- [x] Memory Hierarchy (S0.5) added
+- [x] Proactive Surfacing (S12) added
+- [x] Answer Gating (S13) added
+- [x] Library Protocol (S14) added
+- [x] AI Behavior (S17) added
+- [x] Areeb's 4-step debugging framework integrated into Coding (S4)
+- [x] //turbo mode added
+
+### Phase 11.3: Critical Workflow Rewrites (COMPLETE)
+- [x] `/log` V4 (absorbs /checkup, verbose session logs, CISC)
+- [x] `/analyse-day` V4 (absorbs /context, variable interrogation, scientific patterns)
+- [x] `/plan-tomorrow` V4 (zombie audit, dependency awareness, project scope)
+- [x] `/review-week` V4 (system improvement point, monthly mode, sample size)
+- [x] `/overview` V4 (NEW - absorbs /project-notes, staleness detection, portfolio mode)
+
+### Phase 11.4: Remaining Workflow Rewrites (COMPLETE)
+- [x] `/primer` V4 (session primer, pattern check, energy-based)
+- [x] `/harvest` V4 (mandatory Library dedup, MOC genesis, upward linking)
+- [x] `/context` V4 (DEPRECATED - merged into /analyse-day)
+- [x] `/ignite` V4 (single persona, Library pre-search, leverage audit)
+- [x] `/setup` V4 (Bank G Vitality, Bank H Faith/Purpose, Goals.md, Perpetual Protocol)
+- [x] `/walkthrough` V4 (14 workflows, V4 features, three domains)
+- [x] `/init-plan` V4 (dynamic logic, stack-aware, CISC)
+- [x] `/debug-resistance` V4 (multi-step interrogation, Auto_Learnings integration)
+- [x] `/pivot-strategy` V4 (sample size validation, HIGH-RISK gate, version tracking)
+- [x] `/reboot` V4 (staleness tiers, recurring reboot detection, escalation)
+- [x] `/refactor-plan` V4 (system-triggerable, surgical edit, CISC)
+- [x] `/project-notes` V4 (DEPRECATED - merged into /overview)
+
+### Phase 11.4.1: V4 Compliance Fixes (COMPLETE)
+- [x] Tested all 15 workflows against V4 checklist (13 items) + 11 principles
+- [x] Fixed all 15 workflows: added self-improvement, proactive suggestions, answer gating, CISC, bottom bookend
+- [x] Fixed broken `[V4_ARCHITECTURE]` references (5 occurrences across 4 workflows)
+- [x] Fixed memory hierarchy gaps (reboot P0 order, overview P3/P5, init-plan P0, refactor-plan P0)
+- [x] Fixed Dossier reference in debug-resistance.md
+- [x] Added workflow registry (15 active + 2 deprecated) to Global Rules
+- [x] Added supporting flow paths, three domains, paradigm statement to Global Rules
+- [x] Daily Note Template V4: added `## Session Log` + `## Metrics` sections
+- [x] Created `Goals.md` template in Engine
+- [x] Created `Metrics_Log.md` template in Engine
+- [x] Removed Co-Authored-By attributions from all V4 commits
+- [x] Disabled future attribution in Claude Code settings
+
+### Phase 11.5: Golden Templates (PENDING)
+- [ ] `Golden_Templates/` folder
+- [ ] `Golden_AnalyseDay.json`
+- [ ] `Golden_PlanTomorrow.json`
+- [ ] `Golden_ReviewWeek.json`
+
+### Phase 11.6: Benchmarking System (PENDING)
+- [ ] `V4_Benchmark_Protocol.md`
+- [ ] Sample test data
+- [ ] LLM-as-Judge prompt
+
+### Phase 11.7a: Claude Code Hooks (COMPLETE)
+- [x] Stop hook: session-log-reminder (reminds /log after 5+ exchanges)
+- [x] PreToolUse hook: engine-protection (confirms before writing core Engine files)
+- [x] `.claude/settings.json` with hooks config
+- [x] `.claude/hooks/` directory with executable scripts
+- [x] No jq dependency (pure grep/sed for Windows compatibility)
+
+### Phase 11.7b: Advanced Features (PARTIAL)
+- [x] `/history` workflow (git-based Engine file tracking) — COMPLETE
+- [ ] MCP integration (Health Connect, Strong App, Context7, Chrome)
+- [ ] Skills architecture for large workflows (/ignite, /setup)
+- [ ] Session quality auto-evaluation hook
+
+---
+
+## 5. V4 WORKFLOW PATTERNS (All workflows now follow these)
+
+1. **Pre-flight:** Memory hierarchy load (P0-P5), rules confirmation
+2. **Prerequisite check:** Verify required files exist, chain prerequisites
+3. **Context injection:** Grounded data extraction from Engine files
+4. **Execution:** All claims cite [file:line] or [INFERRED: reason]
+5. **CISC self-verification:** 90%+ accuracy check before output
+6. **Risk-tiered verification:** HIGH=user confirm, MEDIUM=CISC, LOW=direct
+7. **Self-improvement:** Log corrections to Auto_Learnings.md
+8. **Proactive surfacing:** Pattern-based suggestions with confidence scores
+9. **Answer gating:** No handoff until all questions answered
+10. **Handoff (bottom bookend):** Sources, confidence range, next workflow
+
+---
+
+## 6. RULES FOR FUTURE AI INSTANCES
+
+1. **Read this file first** before any workflow changes
+2. **Paradigm:** We do Workflow Engineering, not Prompt Engineering
+3. **Never recreate Mercenary Mode** - deliberately removed
+4. **Confidence scores:** 1-5 scale, threshold 3
+5. **Citations:** [file:line] or [INFERRED: reason] required
+6. **Update progress** in Section 4 after completing tasks
+7. **Append-only:** Never delete from plan.md, project_state.md, or Auto_Learnings.md
+8. **Three domains:** Identity covers Work, Vitality, Faith/Purpose
+9. **Areeb framework:** Self-Critique > Verbose Logging > 5-Why > Differential Diagnosis
+10. **Windows:** Never use && in commands
+
+---
+
+## 7. FILES CREATED/MODIFIED IN V4
+
+| File | Path | Status |
+|------|------|--------|
+| Mover_Global_Rules.md | `src/system/` | Rewritten to V4 |
+| V4_CONTEXT.md | `src/system/` | Updated |
+| log.md | `src/workflows/` | Rewritten to V4 |
+| analyse-day.md | `src/workflows/` | Rewritten to V4 |
+| plan-tomorrow.md | `src/workflows/` | Rewritten to V4 |
+| review-week.md | `src/workflows/` | Rewritten to V4 |
+| overview.md | `src/workflows/` | NEW (V4) |
+| morning.md | `src/workflows/` | Rewritten to V4 |
+| harvest.md | `src/workflows/` | Rewritten to V4 |
+| context.md | `src/workflows/` | DEPRECATED |
+| ignite.md | `src/workflows/` | Rewritten to V4 |
+| setup.md | `src/workflows/` | Rewritten to V4 |
+| walkthrough.md | `src/workflows/` | Rewritten to V4 |
+| init-plan.md | `src/workflows/` | Rewritten to V4 |
+| debug-resistance.md | `src/workflows/` | Rewritten to V4 |
+| pivot-strategy.md | `src/workflows/` | Rewritten to V4 |
+| reboot.md | `src/workflows/` | Rewritten to V4 |
+| refactor-plan.md | `src/workflows/` | Rewritten to V4 |
+| project-notes.md | `src/workflows/` | DEPRECATED |
+
+---
+
+## 8. V4.x INCREMENTAL UPDATES (Post-Launch)
+
+Changes since initial V4 release (2026-02-19):
+
+| Version | Date | Changes |
+|---------|------|---------|
+| 4.0.9 | 2026-02-21 | All 17 workflows intelligence-fixed, workflow compliance audit |
+| 4.0.10 | 2026-02-22 | 13-agent installer (Codex split, Gemini CLI), skill category multi-select, tiered installs |
+| 4.0.11 | 2026-02-23 | Global Rules trimmed (agent-agnostic skills, removed VoltAgent/CHANGELOG/Areeb refs) |
+| 4.0.12 | 2026-02-23 | Hook double-fire fix, installer deep-merge for hooks, tiered workflow reads, Identity Snapshot + Active Patterns in Active_Context |
+| 4.0.13 | 2026-02-23 | Install.js TDZ crash fix, S0 Active_Context read mandate, /update detection + migration handlers |
+| 4.0.14 | 2026-02-23 | /update V3→V4 migration audit (4 detection sections, 7 migration handlers) |
+| 4.0.15 | 2026-02-23 | Correction lifecycle (S11 rewrite, Safe Modification Protocol), /update What's New section, project_state cleanup |
+| 4.0.16 | 2026-02-23 | Tier reclassification (Recommended/Supported/Compatible/Basic), interactive walkthrough rewrite, full documentation update |
+
+### Key Changes Since Initial V4
+
+- **13 agents** (was 11): Codex split into CLI + Cloud, Gemini CLI added
+- **`/primer` renamed to `/morning`** — works the same, better name
+- **`/history` workflow** complete (was PENDING)
+- **`/debrief` workflow** added (post-meeting capture)
+- **38 skills** (was 40+ — consolidated during audit)
+- **Install command:** `npx moveros` (was `npx mover-os`)
+- **Tier labels removed** — table shows Agent + What Gets Installed (no tier column). Claude Code callout for deepest integration.
+- **Correction lifecycle:** S11 Trigger A provides same-session inline fix path with diff format
+- **Safe Modification Protocol:** 7 guardrails for self-editing
+- **Interactive walkthrough:** STOP gates, live demos, capability groups
+- **Missing workflows in file table:** /debrief, /history, /update (added below)
+
+### Additional Files (not in original Section 7)
+
+| File | Path | Status |
+|------|------|--------|
+| debrief.md | `src/workflows/` | NEW (V4) |
+| history.md | `src/workflows/` | NEW (V4) |
+| update.md | `src/workflows/` | NEW (V4) |
+
+---
+
+**END OF V4 CONTEXT FILE**
+*Update after each implementation session.*