autoworkflow 1.1.0 → 2.0.0

package/.claude/commands/fix.md

@@ -0,0 +1,119 @@
+# /fix - Fix Command
+
+Fix verification errors and re-verify.
+
+## Trigger
+- User invokes `/fix`
+- Or: Automatically when `verify_loop` detects errors
+- Or: When `audit_loop` detects issues
+
+## Workflow
+
+### Step 1: Identify Errors
+Parse the error output from verification:
+```
+1. Collect all TypeScript errors
+2. Collect all ESLint warnings
+3. Sort by severity (errors first)
+4. Sort by file (group related errors)
+```
+
+### Step 2: Fix Each Error
+For each error:
+```
+1. Read the file and error context
+2. Understand the issue
+3. Determine the fix
+4. Apply the fix
+5. Mark as addressed
+```
+
+### Step 3: Report Progress
+
+**Output Format:**
+```
+## Fix Loop (Iteration X/10)
+
+Fixing X errors...
+
+### TypeScript Errors
+
+[1/3] `src/Component.tsx:15`
+**Error:** Property 'onClick' is missing in type
+**Fix:** Added onClick to props interface
+**Status:** ✅ Fixed
+
+[2/3] `src/hooks/useAuth.ts:42`
+**Error:** Type 'string' is not assignable to type 'number'
+**Fix:** Changed return type to string
+**Status:** ✅ Fixed
+
+### ESLint Warnings
+
+[3/3] `src/pages/Home.tsx:8`
+**Warning:** 'unused' is defined but never used
+**Fix:** Removed unused import
+**Status:** ✅ Fixed
+
+---
+
+All errors addressed. Re-running verification...
+```
+
+### Step 4: Re-verify
+After fixing, automatically run `/verify` again.
+
+### Step 5: Handle Stuck Errors
+If an error cannot be fixed:
+```
+## Fix Loop
+
+⚠️ **Unable to Fix**
+
+**Error:** [description]
+**File:** `path/to/file.tsx:line`
+**Reason:** [why auto-fix failed]
+
+**Options:**
+1. I can attempt a different approach
+2. You can provide guidance
+3. Skip this error and continue
+
+**What would you like me to do?**
+```
+
+## Fix Strategies
+
+### TypeScript Errors
+| Error Type | Strategy |
+|------------|----------|
+| Missing property | Add to interface |
+| Type mismatch | Adjust type or value |
+| Undefined variable | Add declaration or import |
+| Missing return | Add return statement |
+| Null reference | Add null check |
+
+### ESLint Errors
+| Error Type | Strategy |
+|------------|----------|
+| Unused variable | Remove or use |
+| Missing semicolon | Run lint:fix |
+| console.log | Remove statement |
+| any type | Add proper type |
+| Complexity | Extract function |
+
+## Loop Reference
+See `system/loops.md#fix_loop` for loop definition.
+
+## Max Iterations
+Combined with verify_loop: 10 total iterations.
+
+## Rules
+- Fix ALL errors, not just some
+- Do not stop until verification passes
+- Show what was changed
+- Return to verify_loop after fixes
+
+## Examples
+- `/fix` - Fix current verification errors
+- Automatic when verify fails

package/.claude/commands/plan.md

@@ -0,0 +1,101 @@
+# /plan - Plan Command
+
+Create an implementation plan with suggestions.
+
+## Trigger
+- User invokes `/plan [task]`
+- Or: After ANALYZE phase completes
+
+## Workflow
+
+### Step 1: Verify Analysis Complete
+Check that `analyze_gate` has passed:
+- Relevant files have been read
+- Task has been classified
+- Requirements understood
+
+### Step 2: Design Implementation
+```
+1. Break task into specific changes
+2. List files to create/modify
+3. Identify the order of changes
+4. Note potential issues
+```
+
+### Step 3: Prepare Suggestions
+Based on task type (from `system/router.md`):
+
+| Task Type | Suggestions |
+|-----------|-------------|
+| feature | Required |
+| fix | Conditional |
+| refactor | Conditional |
+| style | Skip |
+| docs | Skip |
+
+### Step 4: Present Plan
+
+**Output Format:**
+```
+## Plan: [task description]
+
+### Task Classification
+**Type:** [from router]
+**Workflow:** [phases]
+
+### Proposed Changes
+1. **[Change 1]**
+   - File: `path/to/file.tsx`
+   - Action: [create/modify/delete]
+   - Details: [what changes]
+
+2. **[Change 2]**
+   - File: `path/to/file.ts`
+   - Action: [create/modify]
+   - Details: [what changes]
+
+### Suggestions for Feature Completeness
+
+🔴 **Required** (must implement):
+- [ ] [Item 1] - [reason]
+- [ ] [Item 2] - [reason]
+
+🟡 **Recommended** (improves quality):
+- [ ] [Item 3] - [reason]
+- [ ] [Item 4] - [reason]
+
+🟢 **Optional** (nice to have):
+- [ ] [Item 5] - [reason]
+
+### Potential Issues
+- [Issue 1 and mitigation]
+- [Issue 2 and mitigation]
+
+---
+
+**Should I proceed with this plan?**
+**Which suggestions to include?** (all/required/none/1,2,4)
+```
+
+## Gate
+`plan_approval_gate` - MUST wait for user approval before implementing.
+
+### Approval Detection
+**Approved:**
+- "yes", "y", "proceed", "go ahead"
+- "approved", "lgtm", "looks good"
+- "all" (all suggestions)
+- "required" (required only)
+- Numbers: "1, 3, 5"
+
+**Not Approved:**
+- "no", "wait", "hold on"
+- Questions or revision requests
+
+## Arguments
+`$ARGUMENTS` - The task to plan (optional if following analysis)
+
+## Examples
+- `/plan add user profile page` - Plan a new feature
+- `/plan fix login error` - Plan a bug fix
+- `/plan` - Continue from analysis to planning

package/.claude/commands/suggest.md

@@ -0,0 +1,195 @@
+# /suggest - Suggest Command
+
+Generate feature completeness suggestions for a task.
+
+## Trigger
+- User invokes `/suggest [task]`
+- Or: Automatically during PLAN phase for features
+
+## When to Show Suggestions
+
+Per `system/router.md`:
+
+| Task Type | Show Suggestions |
+|-----------|------------------|
+| feature | Always |
+| component | Always |
+| page/route | Always |
+| fix | Only if reveals missing handling |
+| refactor | Only if scope expands |
+| style | Skip |
+| config | Skip |
+| docs | Skip |
+
+## Workflow
+
+### Step 1: Analyze Task
+```
+1. Understand what is being built
+2. Check instructions/AI_RULES.md for standards
+3. Check instructions/BLUEPRINT.md for requirements
+4. Identify completeness gaps
+```
+
+### Step 2: Categorize Suggestions
+
+**🔴 Required (must implement):**
+- Error handling for failure cases
+- Loading states for async operations
+- Input validation for user data
+- Accessibility basics (aria labels)
+
+**🟡 Recommended (improves quality):**
+- Empty states when no data
+- Edge case handling
+- Keyboard navigation
+- Better UX feedback
+
+**🟢 Optional (nice to have):**
+- Unit tests
+- Analytics tracking
+- Performance optimizations
+- Additional features
+
+### Step 3: Present Suggestions
+
+**Output Format:**
+```
+## Suggestions: [task]
+
+Based on project standards and the task requirements:
+
+### 🔴 Required (must implement)
+
+1. **Error Handling**
+   - Add error state UI for API failures
+   - Show user-friendly error messages
+
+2. **Loading State**
+   - Add loading skeleton while fetching
+   - Disable submit during processing
+
+3. **Input Validation**
+   - Validate required fields
+   - Show validation errors inline
+
+### 🟡 Recommended (improves quality)
+
+4. **Empty State**
+   - Show message when no results
+   - Suggest actions user can take
+
+5. **Keyboard Support**
+   - Handle Enter to submit
+   - Support Tab navigation
+
+6. **Mobile Layout**
+   - Responsive design
+   - Touch-friendly targets
+
+### 🟢 Optional (nice to have)
+
+7. **Unit Tests**
+   - Test component rendering
+   - Test hook behavior
+
+8. **Performance**
+   - Memoize expensive calculations
+   - Lazy load if large
+
+---
+
+**Which suggestions should I include?**
+- `all` - Include all suggestions
+- `required` - Only required items
+- `none` - Skip suggestions
+- `1, 2, 4, 7` - Specific numbers
+```
+
+## Suggestion Templates
+
+### For New Pages
+```
+🔴 Required:
+- [ ] Loading skeleton
+- [ ] Error boundary
+- [ ] Empty state
+- [ ] Mobile layout
+
+🟡 Recommended:
+- [ ] SEO meta tags
+- [ ] Breadcrumb navigation
+- [ ] Page title
+
+🟢 Optional:
+- [ ] Analytics tracking
+- [ ] Keyboard shortcuts
+```
+
+### For New Forms
+```
+🔴 Required:
+- [ ] Client-side validation
+- [ ] Server error display
+- [ ] Submit loading state
+- [ ] Success feedback
+
+🟡 Recommended:
+- [ ] Autofocus first field
+- [ ] Clear form on success
+- [ ] Confirm unsaved changes
+
+🟢 Optional:
+- [ ] Auto-save draft
+- [ ] Undo/redo support
+```
+
+### For New Components
+```
+🔴 Required:
+- [ ] TypeScript props interface
+- [ ] Handle loading prop
+- [ ] Handle error prop
+
+🟡 Recommended:
+- [ ] Default props
+- [ ] Ref forwarding
+- [ ] Accessibility
+
+🟢 Optional:
+- [ ] Storybook story
+- [ ] Unit tests
+```
+
+### For API Integration
+```
+🔴 Required:
+- [ ] Error handling
+- [ ] Loading state
+- [ ] Type definitions
+
+🟡 Recommended:
+- [ ] Retry logic
+- [ ] Cache response
+- [ ] Cancel on unmount
+
+🟢 Optional:
+- [ ] Optimistic updates
+- [ ] Background refresh
+```
+
+## Key Principle
+
+**Always reference instructions/AI_RULES.md** when making suggestions:
+1. Align with project vision
+2. Meet quality standards
+3. Follow ALWAYS/NEVER rules
+4. Complete feature requirements
+
+## Arguments
+`$ARGUMENTS` - The task to generate suggestions for
+
+## Examples
+- `/suggest add user profile page` - Suggestions for new page
+- `/suggest add contact form` - Suggestions for new form
+- `/suggest` - Suggestions for current task

package/.claude/commands/verify.md

@@ -0,0 +1,113 @@
+# /verify - Verify Command
+
+Run TypeScript and ESLint verification with fix loop.
+
+## Trigger
+- User invokes `/verify`
+- Or: Automatically after IMPLEMENT phase
+- Or: Part of verify_loop
+
+## Workflow
+
+### Step 1: Run Verification
+```bash
+npm run verify
+# Equivalent to: npm run typecheck && npm run lint
+```
+
+### Step 2: Parse Results
+Check exit code and output:
+- Exit 0 = All checks passed
+- Exit non-0 = Errors exist
+
+### Step 3: Report Results
+
+**On Success:**
+```
+## Verification
+
+✅ **PASSED**
+
+| Check | Status | Details |
+|-------|--------|---------|
+| TypeScript | ✅ | 0 errors |
+| ESLint | ✅ | 0 warnings |
+
+**Gate Status:** `verify_gate` ✅ PASSED
+
+Ready to proceed to [AUDIT/COMMIT].
+```
+
+**On Failure:**
+```
+## Verification
+
+❌ **FAILED** (Iteration X/10)
+
+| Check | Status | Details |
+|-------|--------|---------|
+| TypeScript | ⛔ | X errors |
+| ESLint | ⛔ | X warnings |
+
+### Errors Found
+
+**TypeScript Errors:**
+1. `src/Component.tsx:15`
+   Error: Property 'onClick' is missing in type...
+
+2. `src/hooks/useAuth.ts:42`
+   Error: Type 'string' is not assignable to type 'number'
+
+**ESLint Warnings:**
+1. `src/pages/Home.tsx:8`
+   Warning: 'unused' is defined but never used
+
+**Gate Status:** `verify_gate` ⛔ BLOCKED
+
+Entering fix loop...
+```
+
+### Step 4: Fix Loop (if errors)
+Per `system/loops.md#fix_loop`:
+1. Parse each error
+2. Fix one by one
+3. Re-run verification
+4. Repeat until pass or max iterations
+
+### Step 5: Max Iterations
+If max iterations (10) reached:
+```
+## Verification
+
+⛔ **MAX ITERATIONS REACHED** (10/10)
+
+Unable to automatically fix all errors.
+
+### Remaining Errors
+1. [Error 1]
+2. [Error 2]
+
+### What Was Tried
+- [Fix attempt 1]
+- [Fix attempt 2]
+
+**Need your guidance to proceed.**
+```
+
+## Gate
+`verify_gate` - Must pass before AUDIT or COMMIT phase.
+
+## Commands Used
+```bash
+npm run verify      # Full verification
+npm run typecheck   # TypeScript only
+npm run lint        # ESLint only
+npm run lint:fix    # Auto-fix ESLint issues
+```
+
+## Loop Reference
+See `system/loops.md#verify_loop` for loop definition.
+
+## Examples
+- `/verify` - Run verification after making changes
+- Automatic after any implementation

package/.claude/settings.json

@@ -0,0 +1,129 @@
+{
+  "$schema": "https://raw.githubusercontent.com/anthropics/claude-code/main/schemas/settings.json",
+  "version": "1.0.0",
+  "name": "AutoWorkflow - Full Architecture",
+  "description": "System prompt layer that controls Claude's workflow with triggers, loops, and gates",
+
+  "instructions": [
+    "ALWAYS read CLAUDE.md (entry point) at the start of EVERY task",
+    "ALWAYS check if instructions/BLUEPRINT.md exists at session start",
+    "IF BLUEPRINT.md missing: run /audit project to generate it",
+    "ALWAYS check system/router.md to classify the task type",
+    "ALWAYS check system/gates.md before phase transitions",
+    "ALWAYS follow system/loops.md for verification and fix cycles",
+    "ALWAYS check system/triggers.md for event handling",
+    "ALWAYS update BLUEPRINT.md after adding new features/routes/APIs",
+    "NEVER skip the plan approval gate - wait for user confirmation",
+    "NEVER commit without passing all gates in system/gates.md",
+    "NEVER implement without reading instructions/AI_RULES.md",
+    "NEVER build backend without corresponding UI (UI enforcement)",
+    "Git hooks (hooks/pre-commit, hooks/commit-msg) provide backup enforcement",
+    "Scripts in scripts/ folder automate verification and audits"
+  ],
+
+  "requiredFiles": [
+    "CLAUDE.md",
+    "system/triggers.md",
+    "system/loops.md",
+    "system/gates.md",
+    "system/router.md",
+    "instructions/CLAUDE.md",
+    "instructions/AI_RULES.md",
+    "instructions/BLUEPRINT.md"
+  ],
+
+  "automationFiles": [
+    "scripts/setup.sh",
+    "scripts/autoworkflow.sh",
+    "scripts/check-ui-enforcement.sh",
+    "hooks/pre-commit",
+    "hooks/commit-msg"
+  ],
+
+  "workflow": {
+    "mode": "agentic",
+    "phases": ["ANALYZE", "PLAN", "CONFIRM", "IMPLEMENT", "VERIFY", "FIX", "AUDIT", "COMMIT", "UPDATE"],
+    "requirePlanApproval": true,
+    "requireVerifyBeforeCommit": true,
+    "requireAuditForFeatures": true,
+    "requireBlueprintCheck": true,
+    "maxFixIterations": 10,
+    "maxAuditIterations": 5
+  },
+
+  "gates": {
+    "analyze": {
+      "requires": ["Read relevant source files", "Check BLUEPRINT.md"]
+    },
+    "plan_approval": {
+      "requires": ["User explicit approval"],
+      "blocking": true
+    },
+    "verify": {
+      "requires": ["npm run verify passes"],
+      "blocking": true
+    },
+    "audit": {
+      "requires": ["npm run audit:ui passes", "npm run audit:cycles passes"],
+      "blocking": true,
+      "appliesTo": ["feature", "refactor", "perf"]
+    },
+    "pre_commit": {
+      "requires": [
+        "verify_gate passed",
+        "audit_gate passed (if applicable)",
+        "No TODO/FIXME in diff",
+        "No console.log in diff",
+        "Valid commit message format"
+      ],
+      "blocking": true
+    }
+  },
+
+  "triggers": {
+    "on_conversation_start": ["Read CLAUDE.md", "Check BLUEPRINT.md exists", "Route via system/router.md"],
+    "on_blueprint_missing": ["Run project audit", "Generate AI_RULES.md + BLUEPRINT.md", "Present for approval"],
+    "on_task_start": ["Read CLAUDE.md", "Route via system/router.md"],
+    "on_implementation_complete": ["Enter VERIFY phase", "Run verify_loop"],
+    "on_verification_failed": ["Enter FIX phase", "Run fix_loop"],
+    "on_verification_passed": ["Check if audit required", "Proceed accordingly"],
+    "on_commit_requested": ["Run pre_commit_gate", "Wait for approval"],
+    "on_feature_complete": ["Check if BLUEPRINT.md needs update", "Present update for approval"]
+  },
+
+  "loops": {
+    "verify_loop": {
+      "command": "npm run verify",
+      "maxIterations": 10,
+      "onFail": "fix_loop"
+    },
+    "fix_loop": {
+      "maxIterations": 10,
+      "onComplete": "verify_loop"
+    },
+    "audit_loop": {
+      "commands": ["npm run audit:ui", "npm run audit:cycles"],
+      "maxIterations": 5
+    }
+  },
+
+  "commands": {
+    "verify": "npm run verify",
+    "typecheck": "npm run typecheck",
+    "lint": "npm run lint",
+    "audit:ui": "npm run audit:ui",
+    "audit:cycles": "npm run audit:cycles",
+    "format": "npm run format"
+  },
+
+  "rules": {
+    "noPartialImplementations": true,
+    "noTodoComments": true,
+    "noConsoleLogs": true,
+    "noOrphanFeatures": true,
+    "noCircularDependencies": true,
+    "conventionalCommits": true,
+    "oneFeatureAtATime": true,
+    "readBeforeWrite": true
+  }
+}

package/.claude/settings.local.json

@@ -0,0 +1,9 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(git push)",
+      "Bash(npm publish:*)",
+      "Bash(npm version:*)"
+    ]
+  }
+}

package/.prettierrc

@@ -0,0 +1,11 @@
+{
+  "semi": false,
+  "singleQuote": true,
+  "tabWidth": 2,
+  "trailingComma": "es5",
+  "printWidth": 100,
+  "bracketSpacing": true,
+  "arrowParens": "avoid",
+  "endOfLine": "lf",
+  "plugins": ["prettier-plugin-tailwindcss"]
+}

package/.vscode/extensions.json

@@ -0,0 +1,27 @@
+{
+  "recommendations": [
+    // Essential for AutoWorkflow
+    "dbaeumer.vscode-eslint",
+    "esbenp.prettier-vscode",
+
+    // TypeScript
+    "ms-vscode.vscode-typescript-next",
+
+    // Tailwind CSS
+    "bradlc.vscode-tailwindcss",
+
+    // React
+    "dsznajder.es7-react-js-snippets",
+
+    // Git
+    "eamodio.gitlens",
+
+    // Code Quality
+    "streetsidesoftware.code-spell-checker",
+    "usernamehw.errorlens",
+
+    // Claude Code (if available)
+    "anthropic.claude-code"
+  ],
+  "unwantedRecommendations": []
+}