specweave 0.3.10 → 0.3.12

package/CLAUDE.md

@@ -101,7 +101,7 @@ src/                            ← SOURCE OF TRUTH (version controlled)
 .specweave/                     ← FRAMEWORK DATA (always present)
 ├── increments/                 ← Feature development
 ├── docs/                       ← Strategic documentation
-└── config.yaml                 ← Configuration
+└── logs/                       ← Logs and execution history
 ```
 
 **Rules**:
@@ -191,7 +191,6 @@ specweave/
 │   ├── docs/
 │   │   ├── internal/           # Strategic docs
 │   │   └── public/             # Published docs
-│   ├── config.yaml             # SpecWeave's config
 │   └── logs/
 │
 ├── tests/
@@ -259,6 +258,28 @@ All AI-generated files MUST go into increment folders:
 - ✅ Clear context (all files for a feature in one place)
 - ✅ No root clutter
 
+### Runtime Artifacts (NOT Source Controlled)
+
+**Logs and ephemeral files** should NEVER be committed:
+
+```
+❌ NEVER COMMIT:
+.specweave/logs/                        # Framework runtime logs
+.specweave/logs/last-hook-fire          # Hook execution timestamps
+.specweave/increments/*/logs/           # Increment execution logs
+.specweave/cache/                       # Temporary cache
+*.tmp                                   # Temporary files
+```
+
+**Why?**
+- ✅ Logs are execution artifacts, not source code
+- ✅ They change on every run (noise in git history)
+- ✅ They cause unnecessary merge conflicts
+- ✅ They bloat the repository over time
+- ✅ They're developer-specific, not shared state
+
+**Enforcement**: `.gitignore` excludes these patterns automatically.
+
 ---
 
 ## Development Workflow

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "specweave",
-  "version": "0.3.10",
+  "version": "0.3.12",
   "description": "Replace vibe coding with spec-driven development. Smart workflow: /specweave inc auto-closes previous, /specweave build auto-resumes, /specweave progress shows status. PM-led planning, 10 agents, 35+ skills. spec-weave.com",
   "type": "module",
   "main": "dist/index.js",

package/src/commands/specweave.done.md

@@ -548,36 +548,10 @@ Check progress: /list-increments
 
 ## Configuration
 
-**File**: `.specweave/config.yaml`
-
-```yaml
-increment_closure:
-  pm_validation:
-    enabled: true           # ← MUST be true
-    strict_mode: true       # Require all 3 gates to pass
-
-  gates:
-    tasks:
-      require_p1_complete: true
-      require_p2_complete: false  # P2 can be deferred
-      allow_scope_transfer: true
-
-    tests:
-      require_all_passing: true
-      min_coverage: 80           # 80% coverage required
-      allow_skipped: false
-
-    documentation:
-      require_claude_md: true
-      require_readme: true
-      require_changelog: true    # For public APIs
-      allow_inline_only: false
-
-  scope_creep:
-    detect: true
-    max_additional_tasks: 10   # Warn if >10 tasks added
-    auto_transfer: true        # Auto-transfer to next increment
-```
+All PM validation settings use sensible defaults:
+- Tasks: P1 required, P2 can be deferred
+- Tests: All tests must pass, 80% coverage required
+- Documentation: CLAUDE.md, README.md, and CHANGELOG.md must be updated
 
 ---
 

package/src/commands/specweave.next.md

@@ -468,27 +468,11 @@ Your choice? [A/B/C] _
 
 ## Configuration
 
-**File**: `.specweave/config.yaml`
-
-```yaml
-workflow:
-  auto_close:
-    enabled: true              # Enable auto-close in /next
-    strict_mode: true          # Require all 3 gates to pass
-    suggest_next: true         # Suggest next work after closure
-
-  pm_validation:
-    gates:
-      tasks:
-        require_p1_complete: true
-        require_p2_complete: false  # P2 can be deferred
-      tests:
-        require_all_passing: true
-        min_coverage: 80
-      documentation:
-        require_claude_md: true
-        require_readme: true
-```
+Auto-close uses sensible defaults:
+- PM validation gates must pass (tasks, tests, docs)
+- P1 tasks required, P2 can be deferred
+- Tests must pass with 80% coverage
+- Documentation must be updated
 
 ---
 

package/src/commands/specweave.validate.md

@@ -22,7 +22,6 @@ You are helping the user validate a SpecWeave increment with optional AI-powered
 - `--quality`: Run AI quality assessment (LLM-as-judge, ~2k tokens, 1-2 minutes)
 - `--export`: Export AI suggestions to tasks.md automatically
 - `--fix`: Auto-fix HIGH priority issues (experimental, requires confirmation)
-- `--always`: Save quality assessment as default in config.yaml
 
 ## Workflow
 
@@ -132,8 +131,7 @@ Action required:
 **Check in this order**:
 
 1. **If `--quality` flag provided**: Run quality assessment (skip prompt)
-2. **Else if `config.yaml` has `validation.quality_judge.always_run: true`**: Run quality assessment (skip prompt)
-3. **Else**: Prompt user
+2. **Else**: Prompt user
 
 **Prompt format** (if needed):
 ```
@@ -730,53 +728,7 @@ fi
 
 ## Configuration
 
-**File**: `.specweave/config.yaml`
-
-```yaml
-validation:
-  enabled: true
-  auto_validate: true
-  severity_threshold: warning
-
-  # Quality judge settings
-  quality_judge:
-    enabled: true
-    always_run: false       # Set to true with --always flag
-    auto_prompt: true       # Prompt user if always_run is false
-    thresholds:
-      excellent: 90
-      good: 80
-      acceptable: 70
-      needs_work: 0
-    dimensions:
-      clarity: true
-      testability: true
-      completeness: true
-      feasibility: true
-      maintainability: true
-      edge_cases: true
-    max_tokens: 2000
-    export_to_tasks: false  # Set to true to auto-export suggestions
-
-  # Auto-fix settings
-  auto_fix:
-    enabled: true
-    require_confirmation: true  # Always ask before applying fixes
-    max_fixes_per_run: 5        # Limit fixes to avoid large changes
-
-  # Report settings
-  reports:
-    save_to: "reports/validation-report.md"
-    format: markdown
-    include_line_numbers: true
-    include_suggestions: true
-    include_history: true
-
-  # Hooks
-  hooks:
-    post_document_save: true
-    pre_implementation: true
-```
+All validation settings use sensible defaults. Quality assessment is prompted each time unless `--quality` flag is used.
 
 ## Related Commands
 

package/src/skills/increment-quality-judge/SKILL.md

@@ -260,17 +260,10 @@ Suggestions: 3 high priority
 [Full report shown above]
 ```
 
-### Example 2: Always-On Mode
-
-```yaml
-# config.yaml
-validation:
-  quality_judge:
-    always_run: true
-```
+### Example 2: Auto-Run Mode
 
 ```bash
-User: "Validate increment 001"
+User: "Validate increment 001 --quality"
 
 increment-quality-judge:
 ✅ Rule-based: 120/120 passed
@@ -439,9 +432,9 @@ Show detailed report
 **When:** User selects "Export to tasks"
 **Then:** 3 tasks added to tasks.md with priority labels
 
-### TC-005: Always-On Mode
-**Given:** config.yaml has always_run: true
-**When:** User requests validation ("validate quality of increment 001")
+### TC-005: Auto-Run Mode
+**Given:** User provides --quality flag
+**When:** User requests validation ("validate increment 001 --quality")
 **Then:** Quality check runs automatically without prompt
 
 ## Best Practices