codecruise 0.1.0

package/config/commands/refine.md

@@ -0,0 +1,323 @@
+---
+name: refine
+description: Upgrade weak roadmap structure to quality standards
+model: sonnet
+tools: Read, Write, Edit, Glob, Grep
+---
+
+# /refine Command
+
+Analyze and improve roadmap structure interactively. Upgrades weak TODOs to include subfeatures, dependencies, file mappings, and rich descriptions.
+
+## When to Use
+
+- After `/init --assess` shows a low score
+- When inheriting a project with flat TODO lists
+- To improve execution efficiency (deps = correct order, files = faster work)
+
+## Usage
+
+```bash
+/refine                              # Refine all roadmap files
+/refine roadmap/phase-01/            # Refine specific phase
+/refine roadmap/phase-01/feature-1.1.yaml  # Refine specific feature
+/refine --dry-run                    # Show suggestions without applying
+/refine --auto                       # Apply obvious improvements without asking
+```
+
+## Process
+
+### 1. Scan Target Files
+
+```bash
+/refine roadmap/phase-02/
+
+Scanning roadmap/phase-02/...
+Found 3 feature files with 47 TODOs
+```
+
+### 2. Analyze Quality
+
+For each file, check:
+
+| Element | Status | Count |
+|---------|--------|-------|
+| Subfeatures | ✗ Missing | 0/3 features |
+| Dependencies | ✗ Missing | 0/47 TODOs |
+| File mappings | ✗ Missing | 0/47 TODOs |
+| Rich descriptions | ◐ Partial | 12/47 TODOs |
+
+**Current score: 38/100**
+
+### 3. Present Improvement Plan
+
+```
+Improvement Plan for roadmap/phase-02/
+
+1. ADD SUBFEATURES (High Impact)
+   Group 47 flat TODOs into logical subfeatures
+   - feature-2.1: 15 TODOs → ~3 subfeatures
+   - feature-2.2: 18 TODOs → ~4 subfeatures
+   - feature-2.3: 14 TODOs → ~3 subfeatures
+
+2. ADD DEPENDENCIES (High Impact)
+   Identify sequential work and add depends_on
+   - Estimate: ~60% of TODOs have implicit dependencies
+
+3. ADD FILE MAPPINGS (Medium Impact)
+   Map TODOs to target source files
+   - Can infer from TODO names and project structure
+
+4. ENRICH DESCRIPTIONS (Low Impact)
+   Expand one-liner descriptions with acceptance criteria
+   - 35 TODOs have weak descriptions
+
+Proceed with refinement? [Y/n]
+```
+
+### 4. Interactive Refinement
+
+#### Step 4a: Subfeatures
+
+```
+═══════════════════════════════════════════════════════════════
+REFINE: feature-2.1-auth.yaml (15 TODOs)
+═══════════════════════════════════════════════════════════════
+
+Current structure (flat):
+  - todo-001: Create user table
+  - todo-002: Create session table
+  - todo-003: Add password hashing
+  - todo-004: Implement login endpoint
+  - todo-005: Implement logout endpoint
+  - todo-006: Add JWT generation
+  - todo-007: Add JWT validation
+  - todo-008: Create auth middleware
+  - todo-009: Add refresh tokens
+  - todo-010: Write login tests
+  - todo-011: Write logout tests
+  - todo-012: Write middleware tests
+  - todo-013: Add rate limiting
+  - todo-014: Add brute force protection
+  - todo-015: Write security tests
+
+Suggested subfeatures:
+  subfeature-2.1a: Database Schema (todo-001, todo-002)
+  subfeature-2.1b: Core Auth (todo-003 to todo-009)
+  subfeature-2.1c: Middleware (todo-008)
+  subfeature-2.1d: Security (todo-013, todo-014)
+  subfeature-2.1e: Tests (todo-010 to todo-012, todo-015)
+
+[a]ccept  [m]odify  [s]kip  [?]explain
+```
+
+#### Step 4b: Dependencies
+
+```
+═══════════════════════════════════════════════════════════════
+REFINE: Dependencies for subfeature-2.1b
+═══════════════════════════════════════════════════════════════
+
+Analyzing TODO sequence...
+
+Suggested dependencies:
+  todo-003 (password hashing)
+    └─ depends_on: [todo-001]  # needs user table
+
+  todo-004 (login endpoint)
+    └─ depends_on: [todo-002, todo-003, todo-006]
+
+  todo-005 (logout endpoint)
+    └─ depends_on: [todo-002, todo-004]
+
+  todo-006 (JWT generation)
+    └─ depends_on: []  # independent utility
+
+  todo-007 (JWT validation)
+    └─ depends_on: [todo-006]
+
+  todo-009 (refresh tokens)
+    └─ depends_on: [todo-006, todo-007]
+
+[a]ccept all  [r]eview each  [s]kip  [?]explain
+```
+
+#### Step 4c: File Mappings
+
+```
+═══════════════════════════════════════════════════════════════
+REFINE: File mappings for subfeature-2.1b
+═══════════════════════════════════════════════════════════════
+
+Scanning project structure...
+Found: src/modules/auth/, src/lib/, tests/
+
+Suggested mappings:
+  todo-003 (password hashing):
+    files:
+      - src/lib/crypto.ts
+      - src/lib/crypto.test.ts
+
+  todo-004 (login endpoint):
+    files:
+      - src/modules/auth/login.controller.ts
+      - src/modules/auth/login.service.ts
+      - tests/auth/login.test.ts
+
+  todo-006 (JWT generation):
+    files:
+      - src/lib/jwt.ts
+      - src/lib/jwt.test.ts
+
+[a]ccept all  [r]eview each  [s]kip
+```
+
+#### Step 4d: Rich Descriptions
+
+```
+═══════════════════════════════════════════════════════════════
+REFINE: Descriptions
+═══════════════════════════════════════════════════════════════
+
+Found 8 TODOs with weak descriptions.
+
+Example - todo-004:
+  Current:  "Implement login endpoint"
+
+  Suggested:
+    description: |
+      Implement POST /auth/login endpoint.
+
+      Request: { email, password }
+      Response: { accessToken, refreshToken, user }
+
+      Acceptance criteria:
+      - Validate email format
+      - Check password against hash
+      - Return JWT tokens on success
+      - Return 401 on invalid credentials
+      - Log failed attempts
+
+[a]ccept  [e]dit  [s]kip  [n]ext
+```
+
+### 5. Apply Changes
+
+```
+═══════════════════════════════════════════════════════════════
+APPLYING REFINEMENTS
+═══════════════════════════════════════════════════════════════
+
+Updating feature-2.1-auth.yaml...
+  ✓ Added 5 subfeatures
+  ✓ Added 12 dependencies
+  ✓ Added 15 file mappings
+  ✓ Enriched 8 descriptions
+
+Updating feature-2.2-user.yaml...
+  ✓ Added 4 subfeatures
+  ✓ Added 14 dependencies
+  ✓ Added 18 file mappings
+  ✓ Enriched 6 descriptions
+
+Updating feature-2.3-profile.yaml...
+  ✓ Added 3 subfeatures
+  ✓ Added 10 dependencies
+  ✓ Added 14 file mappings
+  ✓ Enriched 4 descriptions
+```
+
+### 6. Show Results
+
+```
+═══════════════════════════════════════════════════════════════
+✓ REFINEMENT COMPLETE
+═══════════════════════════════════════════════════════════════
+
+roadmap/phase-02/ upgraded:
+
+  Before          After
+  ──────────────────────────────
+  Score:     38 → 87
+  Subfeatures: 0 → 12
+  Dependencies: 0 → 36
+  File mappings: 0 → 47
+  Rich descriptions: 12 → 47
+
+Files modified:
+  - roadmap/phase-02/feature-2.1-auth.yaml
+  - roadmap/phase-02/feature-2.2-user.yaml
+  - roadmap/phase-02/feature-2.3-profile.yaml
+
+Next: /run phase-02 to start execution
+```
+
+## Options
+
+| Flag | Description |
+|------|-------------|
+| `--dry-run` | Show what would change without modifying |
+| `--auto` | Apply obvious improvements without confirmation |
+| `--subfeatures` | Only add subfeatures |
+| `--deps` | Only add dependencies |
+| `--files` | Only add file mappings |
+| `--descriptions` | Only enrich descriptions |
+| `--score` | Only show current score |
+
+## Examples
+
+```bash
+# Check score without changes
+/refine --score
+
+# See what would be improved
+/refine --dry-run
+
+# Only add dependencies (fastest impact)
+/refine --deps
+
+# Full auto-refinement
+/refine --auto
+
+# Refine specific feature
+/refine roadmap/phase-01/feature-1.2-auth.yaml
+```
+
+## Inference Rules
+
+### Subfeature Grouping
+
+Group TODOs by:
+1. **Prefix patterns**: `*-schema`, `*-service`, `*-controller`, `*-test`
+2. **Semantic clusters**: database, business logic, API, tests
+3. **Size target**: 3-7 TODOs per subfeature (natural checkpoint size)
+
+### Dependency Detection
+
+Infer dependencies from:
+1. **Naming**: "Create X" before "Use X"
+2. **Patterns**: schema → repository → service → controller → tests
+3. **Keywords**: "after", "requires", "based on" in descriptions
+
+### File Mapping
+
+Infer files from:
+1. **Project structure**: Scan `src/`, `lib/`, `packages/`
+2. **Naming conventions**: `{module}.service.ts`, `{module}.test.ts`
+3. **TODO names**: "Create user service" → `user.service.ts`
+
+### Description Enrichment
+
+Expand descriptions with:
+1. **Input/Output**: What goes in, what comes out
+2. **Acceptance criteria**: Bullet points of done conditions
+3. **Edge cases**: Error handling, validation rules
+
+## Quality Bar
+
+After refinement, the structure should have:
+- ✓ Subfeatures grouping related work
+- ✓ Dependencies defining execution order
+- ✓ File mappings for each TODO
+- ✓ Rich descriptions with acceptance criteria
+- ✓ Score ≥ 70/100

package/config/commands/resume.md

@@ -0,0 +1,72 @@
+---
+name: resume
+description: Re-enter OODA loop from paused state
+model: sonnet
+tools: Read, Write, Edit, Bash, Glob, Grep, Task
+---
+
+# /resume
+
+Re-enter OODA loop at OBSERVE phase with fresh environment sensing.
+
+## Usage
+
+```bash
+/resume                       # Re-enter at OBSERVE
+/resume --skip-current        # Skip current TODO, OBSERVE next
+/resume --from todo-1.1b-006  # Jump to specific TODO
+/resume --restart             # Clear partial work, fresh start
+```
+
+## Behavior
+
+| Scenario | Action |
+|----------|--------|
+| Clean pause (after ACT) | OBSERVE current TODO |
+| Dirty pause (mid-ACT) | OBSERVE, assess partial work |
+| Interrupted session | Drift detection, then OBSERVE |
+| `--restart` | Clear partial work, fresh OBSERVE |
+
+## Drift Detection
+
+If paused > 1 hour or environment changed:
+
+```
+▶ RESUMING with drift detection
+
+  Paused: 3 hours ago
+  Changes: 2 new commits on main
+
+  [c] Continue with fresh OBSERVE
+  [r] Re-plan section
+  [a] Abort
+```
+
+## Output
+
+```
+▶ RESUMING OODA Loop
+
+  Re-entry: OBSERVE phase
+  Paused: 1h 30m ago
+  Drift: none detected
+
+  Starting OBSERVE...
+```
+
+## State Transition
+
+```yaml
+# Before
+execution_state:
+  status: paused
+
+# After
+execution_state:
+  status: running
+  resumed_at: {timestamp}
+```
+
+## OODA Integration
+
+Resume always enters at **OBSERVE**, never mid-loop. See `/run` for full OODA documentation.

package/config/commands/retro.md

@@ -0,0 +1,121 @@
+---
+name: retro
+description: Analyze daily conversations to extract learnings and recommend config improvements
+model: opus
+tools: Read, Write, Glob, Grep, Task
+---
+
+# /retro — Daily Retrospective
+
+Analyze conversations to extract insights, identify friction, and recommend improvements.
+
+## Usage
+
+```bash
+/retro                    # Analyze today
+/retro 2026-01-29         # Specific date
+/retro --week             # Last 7 days
+/retro --apply            # Apply recommendations
+/retro --show             # Show pending recommendations
+/retro --dry-run          # Preview without writing
+```
+
+## Storage
+
+```
+~/.claude/learnings/
+├── log.yaml                  # Stats (forever)
+├── daily/{date}.yaml         # Analysis (7 days)
+└── recommendations/
+    ├── pending.yaml          # Actionable (forever)
+    └── applied.yaml          # Audit trail (forever)
+```
+
+## Signal Detection
+
+| Signal | Detection | Indicates |
+|--------|-----------|-----------|
+| Repeated Clarification | Same question 2+ times | Missing context in rules |
+| Long Exchange | >10 exchanges for simple task | Missing command |
+| Error-Retry Cycle | Error + fix attempts | Workflow issue |
+| Manual Workaround | User provides solution | Automation opportunity |
+| Command Failure | Command failed | Command needs fix |
+| Knowledge Gap | Claude asks for existing context | Missing agent loading |
+| Success Pattern | Smooth completion | Pattern to document |
+
+## Recommendation Mapping
+
+| Signal | Type | Target |
+|--------|------|--------|
+| Repeated clarification | Rule update | `rules/{topic}.md` |
+| Long exchange | New command | `commands/{name}.md` |
+| Command failure | Command fix | `commands/{name}.md` |
+| Knowledge gap | Agent update | `agents/` or `skills/` |
+| Success pattern | Documentation | `rules/` or `skills/` |
+
+## Priority Criteria
+
+| Priority | Criteria |
+|----------|----------|
+| P0 | Blocks work, repeated daily, security risk |
+| P1 | Significant friction, affects multiple tasks |
+| P2 | Minor improvement |
+
+## Output Files
+
+**daily/{date}.yaml**:
+```yaml
+date: 2026-01-30
+sessions_analyzed: 5
+total_exchanges: 147
+
+insights:
+  - id: insight-001
+    type: friction
+    signal: repeated_clarification
+    description: "Asked 3x about file organization"
+    recommendation:
+      type: rule_update
+      target: "rules/standards.md"
+      priority: P1
+```
+
+**recommendations/pending.yaml**:
+```yaml
+recommendations:
+  - id: rec-001
+    from_insights: [insight-001]
+    type: rule_update
+    priority: P1
+    target: "rules/standards.md"
+    title: "Add File Organization Standards"
+    implementation: |
+      ## File Organization
+      | Type | Location |
+      |------|----------|
+      | Commands | `commands/` |
+```
+
+## Process
+
+1. **Discover**: Load sessions from index
+2. **Detect**: Scan for signal patterns
+3. **Extract**: Capture evidence with quotes
+4. **Generate**: Map signals to recommendations
+5. **Prioritize**: P0 > P1 > P2
+6. **Output**: Write to learnings/
+7. **Cleanup**: Delete daily files > 7 days old
+
+## Apply Mode
+
+Interactive confirmation for each change:
+- Preview diff
+- Apply or skip
+- Log to applied.yaml
+
+## Quality Bar
+
+- Accurately identifies friction patterns
+- Evidence-based recommendations
+- Maintains audit trail
+- Respects privacy (no PII in learnings)

package/config/commands/retry.md

@@ -0,0 +1,75 @@
+---
+name: retry
+description: Reset TODO and re-enter OODA at OBSERVE
+model: sonnet
+tools: Read, Write, Edit, Bash, Glob, Grep, Task
+---
+
+# /retry
+
+Reset a failed/skipped TODO and re-enter OODA loop at OBSERVE.
+
+## Usage
+
+```bash
+/retry                          # Retry current/last failed
+/retry todo-1.1b-004            # Retry specific TODO
+/retry --fresh                  # Clear partial work too
+/retry --all-failed             # Retry all failed TODOs
+/retry --reset-confidence       # Clear failure history
+```
+
+## Reset Levels
+
+| Flag | Resets | Use When |
+|------|--------|----------|
+| (default) | Status only | Retry with existing work |
+| `--fresh` | Status + partial work | Implementation was wrong |
+| `--reset-confidence` | Status + OODA history | False failure pattern |
+| `--all-failed` | All failed TODOs | Batch retry |
+
+## Output
+
+```
+🔄 RETRY: todo-1.1b-004
+
+  Previous: failed (3 attempts)
+  OODA: confidence recalculating...
+
+  Entering OBSERVE...
+```
+
+### On Success
+
+```
+✓ RETRY SUCCESSFUL
+
+  TODO: todo-1.1b-004
+  Attempt: 4
+  Commit: abc123
+```
+
+### On Failure
+
+```
+❌ RETRY FAILED
+
+  Error: Test assertion failed
+  Recommendation: replan
+
+  [1] Retry again
+  [2] Retry --fresh
+  [3] Skip
+  [4] Replan section
+```
+
+## Retry vs Resume
+
+| Command | Purpose | OODA Entry |
+|---------|---------|------------|
+| `/resume` | Continue from pause | OBSERVE current |
+| `/retry` | Re-attempt failed/skipped | OBSERVE with reset |
+
+## OODA Integration
+
+Retry resets OODA state for the TODO. Previous failures cleared. See `/run` for full OODA documentation.