sqlew 3.9.0 → 4.0.0

package/docs/CONSTRAINT_INTELLIGENCE.md

@@ -0,0 +1,339 @@
+# Constraint Intelligence System (v4.0.0)
+
+**Status:** Production Ready
+**Release Date:** 2025-11-27
+
+---
+
+## Overview
+
+The Constraint Intelligence System provides duplicate detection and similarity-based suggestions for architectural constraints. It complements the [Decision Intelligence System](DECISION_INTELLIGENCE.md) with constraint-specific scoring and workflows.
+
+### Key Features
+
+- **Duplicate Detection**: Two-phase check (exact match, then similarity)
+- **Hybrid Scoring**: Combines text, tags, layer, recency, and priority
+- **Constraint-Specific Scoring**: Optimized for rule-based content
+- **Pre-Creation Validation**: Check before adding new constraints
+
+---
+
+## Accessing Constraint Suggestions
+
+Use `target: "constraint"` with the suggest tool:
+
+```javascript
+// Default is target: "decision"
+suggest({
+  action: "by_tags",
+  target: "constraint",  // Required for constraint suggestions
+  tags: ["api", "performance"]
+})
+```
+
+---
+
+## Scoring Algorithm
+
+### Components (100 points max)
+
+| Component | Max Points | Description |
+|-----------|-----------|-------------|
+| Tag overlap | 40 | 10 per matching tag (max 4 tags) |
+| Layer match | 25 | Exact layer match |
+| Text similarity | 20 | Levenshtein distance calculation |
+| Recency | 10 | Recently updated constraints score higher |
+| Priority | 5 | Higher priority constraints score higher |
+
+### Recency Tiers
+
+| Age | Points |
+|-----|--------|
+| < 30 days | 10 |
+| 30-90 days | 5 |
+| 90-180 days | 2 |
+| > 180 days | 0 |
+
+### Priority Scores
+
+| Priority | Level | Points |
+|----------|-------|--------|
+| 4 | Critical | 5 |
+| 3 | High | 4 |
+| 2 | Medium | 3 |
+| 1 | Low | 2 |
+
+### Thresholds
+
+| Threshold | Value | Description |
+|-----------|-------|-------------|
+| Default min_score | 30 | Minimum score for suggestions |
+| Duplicate threshold | 70 | Score triggering duplicate warning |
+
+---
+
+## Actions
+
+### by_key - Text Pattern Search
+
+Find similar constraints by text pattern (alias for by_context with text only).
+
+```javascript
+suggest({
+  action: "by_key",
+  target: "constraint",
+  text: "API response time",      // Required: text to match
+  layer: "business",              // Optional: filter by layer
+  limit: 5,                       // Optional: max results (default: 5)
+  min_score: 30                   // Optional: minimum score (default: 30)
+})
+```
+
+### by_tags - Tag-Based Discovery
+
+Find constraints by tag overlap (fast).
+
+```javascript
+suggest({
+  action: "by_tags",
+  target: "constraint",
+  tags: ["api", "performance"],   // Required: tags to match
+  layer: "business",              // Optional: filter by layer
+  limit: 5,                       // Optional
+  min_score: 30                   // Optional
+})
+```
+
+### by_context - Hybrid Search
+
+Comprehensive search combining text, tags, layer, and priority.
+
+```javascript
+suggest({
+  action: "by_context",
+  target: "constraint",
+  text: "database query",         // Optional
+  tags: ["sql"],                  // Optional
+  layer: "data",                  // Optional (also used for scoring)
+  priority: 3,                    // Optional (for scoring)
+  limit: 5,                       // Optional
+  min_score: 30                   // Optional
+})
+
+// Note: At least one of text, tags, or layer required
+```
+
+### check_duplicate - Pre-Creation Check
+
+Two-phase duplicate detection before creating constraints.
+
+```javascript
+suggest({
+  action: "check_duplicate",
+  target: "constraint",
+  text: "API response time must be under 100ms",  // Required
+  category: "performance"                          // Optional: filter by category
+})
+
+// Response:
+{
+  is_duplicate: false,
+  match_type: "similar",        // "exact" | "similar" | "none"
+  existing: null,               // Existing constraint (if exact)
+  similar_constraints: [...],   // Array of similar constraints
+  score: 65,                    // Similarity score of top match
+  recommendation: "Similar constraint found..."
+}
+```
+
+---
+
+## Workflows
+
+### Before Creating a Constraint
+
+```javascript
+// Step 1: Check for duplicates
+const result = await suggest({
+  action: "check_duplicate",
+  target: "constraint",
+  text: "API response time must be under 100ms"
+});
+
+// Step 2: Handle result
+if (result.is_duplicate) {
+  // Exact match exists - update instead
+  console.log("Constraint exists:", result.existing);
+} else if (result.match_type === "similar") {
+  // Review similar constraints
+  console.log("Similar constraints:", result.similar_constraints);
+  // Decide: create new or update existing
+} else {
+  // Safe to create new constraint
+  constraint({ action: "add", ... });
+}
+```
+
+### Finding Related Constraints
+
+```javascript
+// Fast: Tag-based discovery
+const byTags = await suggest({
+  action: "by_tags",
+  target: "constraint",
+  tags: ["security", "authentication"]
+});
+
+// Comprehensive: Hybrid search
+const byContext = await suggest({
+  action: "by_context",
+  target: "constraint",
+  text: "password policy",
+  tags: ["security"],
+  layer: "business"
+});
+
+// Review score_breakdown to understand relevance
+byContext.suggestions.forEach(s => {
+  console.log(s.constraint_text, s.score_breakdown);
+});
+```
+
+---
+
+## Response Structure
+
+### Suggestion Object
+
+```typescript
+{
+  id: number,                    // Constraint ID
+  constraint_text: string,       // Full constraint text
+  category: string,              // Constraint category
+  score: number,                 // Total relevance score (0-100)
+  reason: string,                // Human-readable explanation
+  score_breakdown: {
+    tag_overlap: number,
+    layer_match: number,
+    text_similarity: number,
+    recency: number,
+    priority: number
+  },
+  layer?: string,                // Assigned layer
+  tags?: string[],               // Associated tags
+  ts?: number                    // Last updated timestamp
+}
+```
+
+---
+
+## Comparison: Decision vs Constraint Intelligence
+
+| Feature | Decision Intelligence | Constraint Intelligence |
+|---------|----------------------|------------------------|
+| Target | `"decision"` (default) | `"constraint"` |
+| Scoring max | 100 points | 100 points |
+| Key/Text weight | 40 points | 20 points |
+| Tag weight | 30 points | 40 points |
+| Layer weight | 15 points | 25 points |
+| Recency | - | 10 points |
+| Priority | - | 5 points |
+| Auto-update tier | Yes (60+) | No |
+| Hard block tier | Yes (45-59) | No |
+| Duplicate threshold | 60+ auto-update | 70 warning |
+
+### Why Different Scoring?
+
+**Decisions** focus on key patterns and value similarity (choices already made).
+
+**Constraints** focus on tag overlap and layer match (rules that must be enforced).
+
+---
+
+## Best Practices
+
+### 1. Use Consistent Tags
+
+```javascript
+// Good: Consistent taxonomy
+tags: ["api", "performance", "latency"]
+
+// Avoid: Inconsistent naming
+tags: ["API", "perf", "response-time"]
+```
+
+### 2. Check Before Creating
+
+```javascript
+// Always check for duplicates
+suggest({
+  action: "check_duplicate",
+  target: "constraint",
+  text: "Your constraint text"
+})
+```
+
+### 3. Use by_tags for Discovery
+
+```javascript
+// Fast discovery of related constraints
+suggest({
+  action: "by_tags",
+  target: "constraint",
+  tags: ["security"]
+})
+```
+
+### 4. Review score_breakdown
+
+```javascript
+// Understand why a constraint was suggested
+suggestions.forEach(s => {
+  if (s.score_breakdown.tag_overlap > 30) {
+    // High tag overlap - likely related
+  }
+  if (s.score_breakdown.layer_match === 25) {
+    // Same layer - consider consolidation
+  }
+});
+```
+
+---
+
+## Configuration
+
+**File:** `src/constants.ts`
+
+```typescript
+// Constraint suggestion defaults
+export const CONSTRAINT_SUGGEST_DEFAULTS = {
+  MIN_SCORE: 30,
+  LIMIT: 5,
+  DUPLICATE_THRESHOLD: 70,
+} as const;
+
+// Score weights
+export const CONSTRAINT_SCORE_WEIGHTS = {
+  TAG_OVERLAP: 10,      // Per matching tag (max 40)
+  LAYER_MATCH: 25,      // Exact match
+  TEXT_SIMILARITY: 20,  // Max based on Levenshtein
+  RECENCY_MAX: 10,      // Based on age tiers
+  PRIORITY_MAX: 5,      // Based on priority level
+} as const;
+```
+
+---
+
+## Related Documentation
+
+| Document | Content |
+|----------|---------|
+| [DECISION_INTELLIGENCE.md](DECISION_INTELLIGENCE.md) | Decision duplicate detection |
+| [TOOL_REFERENCE.md](TOOL_REFERENCE.md) | Complete parameter reference |
+| [SHARED_CONCEPTS.md](SHARED_CONCEPTS.md) | Layers, priorities, tags |
+| [TOOL_SELECTION.md](TOOL_SELECTION.md) | Constraint vs Decision selection |
+
+---
+
+**Version:** 4.0.0
+**Last Updated:** 2025-11-27

package/docs/DATABASE_AUTH.md

@@ -1,6 +1,6 @@
 # Database Authentication Configuration
 
-This document describes the authentication configuration system for multi-database support in mcp-sqlew v3.7.0+.
+This document describes the authentication configuration system for multi-database support in mcp-sqlew v4.0.0+.
 
 ## Overview
 

package/docs/DECISION_CONTEXT.md

@@ -4,6 +4,8 @@
 
 The **Decision Context** feature allows you to attach rich documentation to architectural and implementation decisions, explaining **WHY** a decision was made, what alternatives were considered, and the trade-offs involved. This goes beyond simple key-value storage to provide deep historical context that helps future developers (both human and AI) understand past reasoning.
 
+> **Preserved in v4.0.0**: This feature is fully preserved in the v4.0.0 schema refactoring. All decision context data (rationale, alternatives, trade-offs, related links) continues to work without changes. The schema uses `v4_decision_context` table with the same structure and functionality.
+
 ---
 
 ## 🆕 Decision Intelligence Integration (v3.9.0)
@@ -582,6 +584,16 @@ Create traceability:
 
 ---
 
+## Version History
+
+| Version | Changes |
+|---------|---------|
+| **v4.0.0** | Feature fully preserved. Schema refactored with `v4_` prefix: `v4_decision_context` table. No API changes. |
+| **v3.9.0** | Added Decision Intelligence System with duplicate detection and automatic historical record consolidation. Added `check_duplicate`, `versions`, and policy-based management. |
+| **v3.2.2** | Initial Release: `add_decision_context` action, `rationale`, `alternatives_considered`, `tradeoffs` fields, `related_task_id` and `related_constraint_id` linking. |
+
+---
+
 ## Token Efficiency
 
 **Decision Context is optional** - only use it for important decisions where future understanding is critical.

package/docs/DECISION_INTELLIGENCE.md

@@ -1,8 +1,9 @@
-# Decision Intelligence System (v3.9.0)
+# Decision Intelligence System (v4.0.0)
 
 **Status:** Production Ready ✅
-**Release Date:** 2025-01-15
+**Release Date:** 2025-11-27
 **Test Coverage:** 495/495 tests passing (100%)
+**v3.9.0 Compatibility:** All Decision Intelligence features from v3.9.0 are preserved and fully functional in v4.0.0
 
 ---
 
@@ -23,7 +24,7 @@ The Decision Intelligence System provides AI-driven duplicate detection, similar
 
 ## Three-Tier Similarity Detection
 
-### Threshold Configuration (v3.9.0)
+### Threshold Configuration (v4.0.0)
 
 ```typescript
 // src/constants.ts
@@ -281,7 +282,7 @@ decision({
 })
 ```
 
-### Behavior (v3.9.0 Three-Tier System)
+### Behavior (v4.0.0 Three-Tier System)
 
 When a decision matches a policy with `suggest_similar=1`:
 1. Similarity detection runs automatically BEFORE decision creation
@@ -432,7 +433,7 @@ suggest({
 
 ## Threshold Adjustment Rationale
 
-**v3.9.0 (2025-11-14):** Three-tier system (45/60 with auto-update)
+**v4.0.0 (2025-11-27):** Three-tier system preserved from v3.9.0 (45/60 with auto-update)
 
 ### Why Three-Tier System?
 
@@ -494,7 +495,7 @@ suggest({
 **File:** `src/constants.ts`
 
 ```typescript
-// Suggestion & Duplicate Detection (v3.9.0 Three-Tier System)
+// Suggestion & Duplicate Detection (v4.0.0 Three-Tier System)
 export const SUGGEST_THRESHOLDS = {
   GENTLE_NUDGE: 35,       // Warning threshold (non-blocking, Tier 1)
   HARD_BLOCK: 45,         // Blocking threshold (forces choice, Tier 2)
@@ -586,9 +587,19 @@ When you hit a hard block:
 
 ---
 
+## Related Documentation
+
+| Document | Content |
+|----------|---------|
+| [CONSTRAINT_INTELLIGENCE.md](CONSTRAINT_INTELLIGENCE.md) | Constraint duplicate detection |
+| [TOOL_REFERENCE.md](TOOL_REFERENCE.md) | Complete parameter reference |
+| [SHARED_CONCEPTS.md](SHARED_CONCEPTS.md) | Layers, priorities, tags |
+
+---
+
 ## Support & Feedback
 
 **Issue Tracker:** https://github.com/anthropics/mcp-sqlew/issues
 **Documentation:** See `/docs` directory
-**Version:** v3.9.0
-**Last Updated:** 2025-11-14
+**Version:** v4.0.0
+**Last Updated:** 2025-11-27

package/docs/GIT_AWARE_AUTO_COMPLETE.md

@@ -1,8 +1,9 @@
-# Git-Aware Auto-Complete (v3.4.0 - v3.5.2)
+# Git-Aware Auto-Complete (v3.4.0 - v4.0.0)
 
 **Feature Status**: Implemented ✅
-**Current Version**: v3.5.2 (Two-Step Git-Aware Workflow)
-**Previous Version**: v3.4.0 (Commit-Based Auto-Complete)
+**Current Version**: v4.0.0 (Two-Step Git-Aware Workflow with v4_ Table Prefix)
+**Previous Version**: v3.5.2 (Two-Step Git-Aware Workflow)
+**Original Version**: v3.4.0 (Commit-Based Auto-Complete)
 **Architecture Decision**: #001 - Replace flawed auto-revert with Git-based task completion
 
 ## Version Evolution
@@ -26,7 +27,7 @@ The current stale task detection system has a critical design flaw:
 // In src/utils/task-stale-detection.ts lines 105-116
 // FLAWED LOGIC: Reverts completed work after 24h idle
 const waitingReviewTransitioned = db.prepare(`
-  UPDATE t_tasks
+  UPDATE v4_tasks
   SET status_id = ?,
       updated_ts = unixepoch()
   WHERE status_id = ?
@@ -137,7 +138,7 @@ git log --since="@1698765432" --name-only -- src/auth.ts
 **AI Time**: 5 minutes
 **Token Budget**: 2k-3k tokens
 
-**File**: `/home/kitayama/TypeScriptProject/mcp-sqlew/src/utils/task-stale-detection.ts`
+**File**: `src/utils/task-stale-detection.ts`
 **Lines to Remove**: 105-116
 
 ```typescript
@@ -171,7 +172,7 @@ totalTransitioned += waitingReviewTransitioned.changes;
 **Token Budget**: 18k-25k tokens
 
 **New Function**: `detectAndCompleteReviewedTasks(db: Database): Promise<number>`
-**Location**: `/home/kitayama/TypeScriptProject/mcp-sqlew/src/utils/task-stale-detection.ts`
+**Location**: `src/utils/task-stale-detection.ts`
 
 **Function Signature**:
 ```typescript
@@ -197,7 +198,7 @@ export async function detectAndCompleteReviewedTasks(db: Database): Promise<numb
 1. Check if git_auto_complete_enabled config is true
 2. Get all tasks in waiting_review status
 3. For each task:
-   a. Get list of watched files from t_task_file_links
+   a. Get list of watched files from v4_task_file_links
    b. Get task.created_ts
    c. For each watched file:
       - Run: git log --since="@<created_ts>" --name-only -- <file_path>
@@ -230,7 +231,7 @@ export async function detectAndCompleteReviewedTasks(db: Database): Promise<numb
 **AI Time**: 10-15 minutes
 **Token Budget**: 8k-12k tokens
 
-**Files**: `/home/kitayama/TypeScriptProject/mcp-sqlew/src/tools/tasks.ts`
+**Files**: `src/tools/tasks.ts`
 **Actions to Modify**: `list` (line ~750), `stats` in utils.ts
 
 **Changes**:
@@ -275,8 +276,8 @@ async function listTasks(db: Database, params: ListTasksParams): Promise<object>
 **Token Budget**: 10k-15k tokens
 
 **Files**:
-- `/home/kitayama/TypeScriptProject/mcp-sqlew/src/watcher/file-watcher.ts`
-- `/home/kitayama/TypeScriptProject/mcp-sqlew/src/watcher/index.ts`
+- `src/watcher/file-watcher.ts`
+- `src/watcher/index.ts`
 
 **Changes**:
 1. Watch `.git/index` file for changes (indicates git commit)
@@ -327,7 +328,7 @@ this.watcher.on('change', async (path) => {
 **AI Time**: 10 minutes
 **Token Budget**: 5k-8k tokens
 
-**File**: `/home/kitayama/TypeScriptProject/mcp-sqlew/src/tools/utils.ts` (stats action)
+**File**: `src/tools/utils.ts` (stats action)
 
 **Changes**:
 Add `review_status` section to stats output showing tasks awaiting Git commits:
@@ -339,8 +340,8 @@ const reviewStatus = db.prepare(`
     COUNT(DISTINCT t.id) as awaiting_commit,
     COUNT(DISTINCT tfl.file_id) as total_files,
     COUNT(DISTINCT CASE WHEN <committed_check> THEN tfl.file_id END) as committed_files
-  FROM t_tasks t
-  JOIN t_task_file_links tfl ON t.id = tfl.task_id
+  FROM v4_tasks t
+  JOIN v4_task_file_links tfl ON t.id = tfl.task_id
   WHERE t.status_id = ?
 `).get(TASK_STATUS.WAITING_REVIEW);
 
@@ -369,8 +370,8 @@ return {
 **Token Budget**: 3k-5k tokens
 
 **Files**:
-- `/home/kitayama/TypeScriptProject/mcp-sqlew/src/database.ts` (config defaults)
-- `/home/kitayama/TypeScriptProject/mcp-sqlew/docs/CONFIGURATION.md`
+- `src/database.ts` (config defaults)
+- `docs/CONFIGURATION.md`
 
 **New Config Keys**:
 ```typescript
@@ -423,12 +424,12 @@ config action=update key="stale_review_notification_hours" value="72"
 **Token Budget**: 23k-32k tokens
 
 **Documentation Files**:
-1. `/home/kitayama/TypeScriptProject/mcp-sqlew/docs/TASK_OVERVIEW.md` - Update lifecycle diagram
-2. `/home/kitayama/TypeScriptProject/mcp-sqlew/docs/AUTO_FILE_TRACKING.md` - Add Git-aware section
-3. `/home/kitayama/TypeScriptProject/mcp-sqlew/docs/TASK_ACTIONS.md` - Update status transitions
-4. `/home/kitayama/TypeScriptProject/mcp-sqlew/CHANGELOG.md` - Add v3.4.0 entry
+1. `docs/TASK_OVERVIEW.md` - Update lifecycle diagram
+2. `docs/AUTO_FILE_TRACKING.md` - Add Git-aware section
+3. `docs/TASK_ACTIONS.md` - Update status transitions
+4. `CHANGELOG.md` - Add v3.4.0 entry
 
-**Test Files** (all in `/home/kitayama/TypeScriptProject/mcp-sqlew/src/tests/`):
+**Test Files** (all in `src/tests/`):
 1. `tasks.git-auto-complete.test.ts` - Comprehensive test suite
 2. `tasks.git-integration.test.ts` - Git log integration tests
 

package/docs/SHARED_CONCEPTS.md

@@ -12,33 +12,57 @@
 
 ## Architecture Layers
 
-5-layer architecture for organizing decisions, constraints, file changes, and tasks:
+9-layer architecture for organizing decisions, constraints, file changes, and tasks:
 
-### 1. **presentation** - User Interface
+### FILE_REQUIRED Layers (6)
+
+#### 1. **presentation** - User Interface
 UI components, API endpoints, CLI handlers, forms, response formatting
 
 **Examples**: React components, REST controllers, web forms
 
-### 2. **business** - Business Logic
+#### 2. **business** - Business Logic
 Core application logic, business rules, domain operations
 
 **Examples**: Service classes, domain models, workflows, validation rules
 
-### 3. **data** - Data Access
+#### 3. **data** - Data Access
 Data persistence and retrieval
 
 **Examples**: Database schemas, repositories, ORMs, queries
 
-### 4. **infrastructure** - Infrastructure
+#### 4. **infrastructure** - Infrastructure
 Technical capabilities and external services
 
 **Examples**: Auth, logging, message queues, caching, email/SMS services
 
-### 5. **cross-cutting** - Cross-Cutting Concerns
+#### 5. **cross-cutting** - Cross-Cutting Concerns
 Aspects spanning multiple layers
 
 **Examples**: Error handling, security, performance, i18n, configuration
 
+#### 6. **documentation** - Documentation & Knowledge
+Project documentation, guides, API docs, code comments, design rationales
+
+**Examples**: README files, API documentation, architecture guides, inline comments
+
+### FILE_OPTIONAL Layers (3)
+
+#### 7. **planning** - Planning & Requirements
+Project planning, roadmap decisions, requirements gathering, estimation
+
+**Examples**: Roadmap items, feature specifications, sprint planning, estimation notes
+
+#### 8. **coordination** - Team Coordination
+Team communication, progress tracking, meeting notes, status updates
+
+**Examples**: Meeting notes, status updates, progress reports, team announcements
+
+#### 9. **review** - Review & Quality Assurance
+Code reviews, testing decisions, quality metrics, approval workflows
+
+**Examples**: Review comments, testing strategies, quality gates, approval notes
+
 ---
 
 ## Enum Values Reference
@@ -51,6 +75,10 @@ type Layer =
   | "data"              // Database, repositories, persistence
   | "infrastructure"    // Auth, logging, external services
   | "cross-cutting"     // Security, error handling, i18n
+  | "documentation"     // Project docs, guides, comments
+  | "planning"          // Planning, roadmap, requirements (file-optional)
+  | "coordination"      // Team communication, status updates (file-optional)
+  | "review"            // Code reviews, QA, approval (file-optional)
 ```
 
 ### status (Decision/Entity Status)
@@ -104,15 +132,19 @@ type TaskStatus =
   | "blocked"          // Cannot proceed due to blocker
   | "done"             // Completed and approved
   | "archived"         // Archived for historical reference
+  | "rejected"         // Rejected/cancelled (v4.1.0)
 ```
 
-**Valid Transitions** (enforced by state machine):
-- `todo` → `in_progress`, `blocked`, `archived`
-- `in_progress` → `waiting_review`, `blocked`, `todo`, `archived`
-- `waiting_review` → `done`, `in_progress`, `blocked`, `archived`
-- `blocked` → `todo`, `in_progress`, `archived`
-- `done` → `archived`
-- `archived` → (terminal state, no transitions)
+**Valid Transitions** (v4.1.0 - relaxed rules):
+
+| Status Type | Statuses | Can Transition To |
+|-------------|----------|-------------------|
+| **Non-terminal** | todo, in_progress, waiting_review, blocked, done | Any status (including terminal) |
+| **Terminal** | archived, rejected | None (final states) |
+
+- Non-terminal statuses can freely move to any other status
+- Terminal statuses (`archived`, `rejected`) cannot transition
+- `rejected` requires optional `rejection_reason` parameter
 
 **Auto-Stale Detection & Auto-Archive**:
 - `in_progress` >2 hours → auto-move to `waiting_review`
@@ -148,7 +180,6 @@ Determines batch operation failure handling:
 
 ### Supported Tools
 - `decision`: `set_batch`
-- `message`: `send_batch`
 - `file`: `record_batch`
 - `task`: `create_batch`
 
@@ -175,7 +206,6 @@ Determines batch operation failure handling:
 ### Available Actions
 
 - **decision**: set, get, list, search_tags, search_layer, versions, set_batch, help
-- **message**: send, get, mark_read, send_batch, help
 - **file**: record, get, check_lock, record_batch, help
 - **constraint**: add, get, deactivate, help
 - **stats**: layer_summary, db_stats, clear, help
@@ -192,4 +222,4 @@ Enum values stored as integers (MCP tools auto-convert - use strings in calls):
 - **msg_type**: 1=decision, 2=warning, 3=request, 4=info
 - **priority**: 1=low, 2=medium, 3=high, 4=critical
 - **change_type**: 1=created, 2=modified, 3=deleted
-- **task_status**: 1=todo, 2=in_progress, 3=waiting_review, 4=blocked, 5=done, 6=archived
+- **task_status**: 1=todo, 2=in_progress, 3=waiting_review, 4=blocked, 5=done, 6=archived, 7=rejected