claude-cli-advanced-starter-pack 1.0.0

package/templates/commands/create-task-list.template.md

@@ -0,0 +1,382 @@
+---
+description: Create intelligent task list with codebase exploration and clarifying questions
+type: project
+complexity: medium
+model: opus
+argument-hint: [task or feature description]
+allowed-tools:
+  - Read
+  - Grep
+  - Glob
+  - Task
+  - TaskCreate
+  - TaskUpdate
+  - TaskList
+  - AskUserQuestion
+  - Write
+---
+
+# /create-task-list - Intelligent Task List Generator
+
+**Create a comprehensive task list with codebase exploration, clarifying questions, and persistent testing instructions.**
+
+---
+
+## QUICK START
+
+```bash
+/create-task-list fix the login page redirect bug
+/create-task-list add dark mode to the settings panel
+/create-task-list
+```
+
+---
+
+## PERSISTENT INSTRUCTIONS
+
+**IMPORTANT**: Reference the persistent testing/deployment rules from `.claude/task-lists/TESTING_RULES.md`.
+
+These rules are loaded automatically and apply to ALL task lists created by this command.
+
+**Quick Reference** (full details in TESTING_RULES.md):
+- **URLs**: {{devEnvironment.tunnel.service}} (`{{urls.tunnel.frontend}}`) or production (`{{urls.production.frontend}}`)
+- **Login**: Use credentials from `{{testing.credentials.usernameEnvVar}}` / `{{testing.credentials.passwordEnvVar}}`
+- **Login Selectors**: `{{testing.selectors.username}}`, `{{testing.selectors.password}}`, `{{testing.selectors.loginButton}}`
+- **Backend**: {{deployment.backend.platform}} (never local dev server)
+- **Deploy workflow**: Delete `/{{frontend.distDir}}` → rebuild → deploy → verify SHA
+- **Task workflow**: Debug with curl → {{testing.e2e.framework}} test → commit after each task
+
+---
+
+## EXECUTION FLOW
+
+### Step 0: Load Persistent Rules
+
+**FIRST ACTION**: Read `.claude/task-lists/TESTING_RULES.md` to load testing and deployment rules.
+
+These rules govern:
+- Which URLs to use ({{devEnvironment.tunnel.service}} vs {{urls.production.frontend}})
+- Login credentials
+- Deployment workflow (delete {{frontend.distDir}}, rebuild, verify SHA)
+- Task completion workflow (curl → {{testing.e2e.framework}} → commit)
+
+---
+
+### Step 1: Capture User Prompt
+
+```
+PROMPT: $ARGUMENTS
+
+If no arguments provided:
+  → Use AskUserQuestion: "What would you like to accomplish today?"
+```
+
+### Step 2: Context Assessment
+
+Evaluate the prompt for:
+
+| Factor | Weight | Indicators |
+|--------|--------|------------|
+| **Specificity** | 30% | File names, function names, error messages |
+| **Scope** | 25% | Single file vs multi-file, frontend vs backend |
+| **Technical depth** | 25% | Stack trace, API endpoints, component names |
+| **Reproducibility** | 20% | Steps to reproduce, expected vs actual behavior |
+
+**Score Calculation**:
+- **70-100%**: Sufficient context → Proceed to Step 3a
+- **0-69%**: Insufficient context → Proceed to Step 3b
+
+---
+
+### Step 3a: SUFFICIENT CONTEXT PATH
+
+Deploy **parallel exploration agents** to understand the issue:
+
+```
+Deploy in SINGLE message (parallel execution):
+
+Task 1: subagent_type="Explore"
+  prompt: "Find files related to: $ARGUMENTS. Identify:
+    - Primary files involved
+    - Related components/services
+    - Recent changes in git history"
+  description: "Explore codebase for context"
+
+Task 2: subagent_type="Explore"
+  prompt: "Search for error patterns, tests, and documentation related to: $ARGUMENTS"
+  description: "Find tests and docs"
+
+Task 3 (if backend-related): subagent_type="Explore"
+  prompt: "Find API endpoints, database models, and service files related to: $ARGUMENTS"
+  description: "Explore backend architecture"
+```
+
+**Wait for all agents to complete**, then proceed to Step 4.
+
+---
+
+### Step 3b: INSUFFICIENT CONTEXT PATH
+
+1. **Quick codebase scan**:
+   ```
+   Deploy Task: subagent_type="Explore"
+     prompt: "Perform a brief codebase overview to understand the project structure
+       and identify areas that might relate to: $ARGUMENTS
+       Also review CLAUDE.md for project-specific patterns."
+     description: "Quick codebase scan"
+   ```
+
+2. **Ask clarifying questions** using AskUserQuestion:
+   - What specific behavior are you seeing?
+   - What is the expected behavior?
+   - Which area of the app is affected (Web UI, Backend API, Desktop)?
+   - Can you provide any error messages or screenshots?
+
+3. **After user provides clarity**, proceed to Step 3a with updated context.
+
+---
+
+### Step 4: Synthesize Findings & Ask Questions
+
+Present findings in a **simple paragraph + bullet point format** (NO CODE):
+
+```markdown
+## Understanding
+
+Based on my exploration, here's what I found:
+
+**The Issue:**
+[1-2 sentence summary of the problem]
+
+**Affected Areas:**
+- [Area 1]: [Brief description]
+- [Area 2]: [Brief description]
+
+**Proposed Approach:**
+1. [Step 1 - What we'll do, not how]
+2. [Step 2]
+3. [Step 3]
+
+**Key Files:**
+- `path/to/file1.ts` - [Why it's relevant]
+- `path/to/file2.py` - [Why it's relevant]
+```
+
+**MANDATORY QUESTIONS** (use AskUserQuestion with multiple questions):
+
+**Question 1: Testing Approach**
+```
+header: "Testing"
+question: "Do you want to use 'Ralph Wiggum Loop' style testing? (Continuous test-fix cycle until all tests pass)"
+options:
+  - label: "Yes - Ralph Loop (Recommended)"
+    description: "Automatically retry fixes until tests pass, max 10 iterations"
+  - label: "No - Manual Testing"
+    description: "I'll run tests manually after each change"
+  - label: "Minimal Testing"
+    description: "Only test at the end of all tasks"
+```
+
+**Question 2: {{testing.e2e.framework}} Environment**
+```
+header: "{{testing.e2e.framework}}"
+question: "Where should {{testing.e2e.framework}} E2E tests run?"
+options:
+  - label: "{{devEnvironment.tunnel.service}} ({{urls.tunnel.frontend}}) (Recommended)"
+    description: "Test against local backend via {{devEnvironment.tunnel.service}} tunnel"
+  - label: "{{urls.production.frontend}} (Production)"
+    description: "Test against production - requires deployment"
+  - label: "No {{testing.e2e.framework}}"
+    description: "Skip E2E tests, use curl/unit tests only"
+```
+
+**Question 3: GitHub Integration**
+```
+header: "GitHub"
+question: "Would you like to create a tracked GitHub issue for this task list?"
+options:
+  - label: "Yes - Create GitHub issue (Recommended)"
+    description: "Creates issue with codebase analysis, adds to project board, auto-updates progress"
+  - label: "No - Local task list only"
+    description: "Just use TodoWrite without GitHub tracking"
+```
+
+**Question 4: Confirm Plan** (only if plan needs adjustment)
+```
+header: "Plan"
+question: "Does this approach look correct, or should I adjust?"
+options:
+  - label: "Looks good - proceed"
+    description: "Create the task list and start working"
+  - label: "Need adjustments"
+    description: "I'll describe what to change"
+```
+
+---
+
+### Step 5: Create Task List
+
+Use **TaskCreate** to build the task list with Claude's native system.
+
+**Task List Structure**:
+
+```
+Task 0: "Review Testing Rules" (DO NOT MARK COMPLETE)
+  description: "Read .claude/task-lists/TESTING_RULES.md before starting. These rules apply to all tasks."
+  activeForm: "Maintaining context"
+
+Task 1: "Login via {{testing.e2e.framework}}" (ALWAYS FIRST for E2E tests)
+  description: |
+    Start {{testing.e2e.framework}} from the LOGIN PAGE. Fresh browser has no session.
+
+    **Use these EXACT selectors (from tech-stack.json):**
+
+    1. Navigate to: {env}/login
+    2. Fill username: `{{testing.selectors.username}}`
+    3. Fill password: `{{testing.selectors.password}}`
+    4. Click: `{{testing.selectors.loginButton}}`
+    5. Wait for: `{{testing.selectors.loginSuccess}}`
+
+    This task is REQUIRED before any other {{testing.e2e.framework}} interaction.
+  activeForm: "Logging into application"
+
+Task 2: "[First actual task]"
+  description: "[What needs to be done]"
+  activeForm: "[Present participle form]"
+
+Task 2: "[Second task]"
+  ...
+
+Task N-1: "Run final verification tests"
+  description: "Execute {{testing.e2e.framework}} E2E tests on {selected_environment} to verify all changes work"
+  activeForm: "Running verification tests"
+
+Task N: "Commit all changes"
+  description: "Create a git commit with all changes from this session"
+  activeForm: "Committing changes"
+```
+
+**IMPORTANT**:
+- Task 0 with persistent instructions should NEVER be marked complete
+- Each task should be atomic and independently verifiable
+- Include curl testing tasks BEFORE {{testing.e2e.framework}} tasks when debugging
+
+---
+
+### Step 6: Create GitHub Issue (If Selected)
+
+If user selected GitHub integration:
+
+1. **Create comprehensive issue** using `/github-create-task`:
+   ```bash
+   /github-create-task --batch-mode
+
+   Title: [Task description from user prompt]
+   Priority: P2-Medium
+   Labels: [Determined from exploration - frontend, backend, feature, bug, etc.]
+   QA: [Required if P0/P1, else Not Required]
+   Stack: [Frontend only|Backend only|Both]
+
+   ## Description
+
+   [User's original prompt]
+
+   ## Codebase Analysis
+
+   [Include findings from Step 3 exploration agents]
+
+   ## Task Checklist
+
+   [Mirror the TodoWrite tasks as checkboxes]
+   - [ ] Task 1: ...
+   - [ ] Task 2: ...
+   ```
+
+2. **Store issue number** for progress tracking:
+   ```json
+   // Write to .claude/task-lists/session-{timestamp}.json
+   {
+     "github_issue": {
+       "number": [ISSUE_NUM],
+       "url": "[ISSUE_URL]",
+       "created_at": "[ISO timestamp]"
+     }
+   }
+   ```
+
+3. **Install progress hook** (auto-updates GitHub issue when tasks complete):
+   - The hook watches for TodoWrite calls
+   - When a task is marked `completed`, it updates the GitHub issue:
+     - Checks off the corresponding checkbox
+     - Adds a progress comment if significant milestone
+   - See `.claude/hooks/tools/github-progress-hook.js`
+
+---
+
+### Step 7: Configure Ralph Loop (If Selected)
+
+If user selected Ralph Loop testing, write configuration:
+
+```javascript
+// Write to .claude/ralph-loop.local.md
+const config = {
+  iteration: 1,
+  max_iterations: 10,
+  completion_promise: "all tasks complete and tests passing",
+  started_at: new Date().toISOString(),
+  task_list_id: "[current session]"
+};
+```
+
+---
+
+### Step 8: Compact Session & Begin
+
+1. **Display task summary**:
+   ```
+   ## Task List Created
+
+   | # | Task | Status |
+   |---|------|--------|
+   | 0 | Persistent Context | (always active) |
+   | 1 | [Task 1 subject] | pending |
+   | 2 | [Task 2 subject] | pending |
+   ...
+
+   **Testing Mode:** [Ralph Loop / Manual / Minimal]
+   **Environment:** [{{devEnvironment.tunnel.service}} / {{urls.production.frontend}}]
+   **GitHub Issue:** [#123 (if created) | Not tracked]
+
+   Ready to begin. Starting with Task 1...
+   ```
+
+2. **Compact context** (use /compact if token usage is high)
+
+3. **Begin Task 1**:
+   - Use `TaskUpdate` to set Task 1 to `in_progress`
+   - Execute the task
+   - Use `TaskUpdate` to set Task 1 to `completed`
+   - Commit changes
+   - Move to Task 2
+
+---
+
+## ERROR HANDLING
+
+| Situation | Action |
+|-----------|--------|
+| Exploration agent fails | Retry with broader search terms |
+| User provides unclear answers | Ask follow-up questions with examples |
+| Task creation fails | Fall back to manual task tracking |
+| Ralph loop exceeds max iterations | Stop loop, summarize progress, ask user how to proceed |
+| {{testing.e2e.framework}} tests fail on {{devEnvironment.tunnel.service}} | Check tunnel status, verify backend is running |
+| {{testing.e2e.framework}} tests fail on production | Verify deployment completed, check SHA match |
+
+---
+
+## RELATED COMMANDS
+
+- `/ralph` - Persistent loop execution
+- `/test-{{testing.e2e.framework}}-{{devEnvironment.tunnel.service}}` - E2E testing with {{devEnvironment.tunnel.service}}
+- `/github-task-start` - Start GitHub Project Board tasks
+- `/create-phase-dev` - Complex multi-phase project planning

package/templates/commands/deploy-full.template.md

@@ -0,0 +1,261 @@
+---
+description: Parallel full-stack deployment (backend + frontend)
+model: haiku
+---
+
+# /deploy-full - Full-Stack Deployment
+
+Deploy both backend and frontend in parallel for fastest deployment.
+
+{{#if deployment.backend.platform}}
+{{#if deployment.frontend.platform}}
+
+## Deployment Targets
+
+| Component | Platform | URL |
+|-----------|----------|-----|
+| Backend | {{deployment.backend.platform}} | {{deployment.backend.productionUrl}} |
+| Frontend | {{deployment.frontend.platform}} | {{deployment.frontend.productionUrl}} |
+
+## Pre-deployment Checklist
+
+```bash
+# 1. Verify clean working directory
+git status
+
+# 2. Ensure on correct branch
+git branch --show-current
+
+# 3. Run tests
+{{#if testing.e2e.testCommand}}
+{{testing.e2e.testCommand}}
+{{else}}
+npm test
+{{/if}}
+```
+
+---
+
+## Backend Deployment
+
+{{#if (eq deployment.backend.platform "railway")}}
+### Railway
+
+Using Railway MCP server for deployment:
+
+```javascript
+// Trigger deployment via Railway MCP
+mcp__railway-mcp-server__deployment_trigger({
+  projectId: "{{deployment.backend.projectId}}",
+  serviceId: "{{deployment.backend.serviceId}}",
+  environmentId: "{{deployment.backend.environmentId}}"
+})
+```
+
+Monitor deployment:
+```javascript
+mcp__railway-mcp-server__deployment_logs({
+  projectId: "{{deployment.backend.projectId}}",
+  serviceId: "{{deployment.backend.serviceId}}"
+})
+```
+{{/if}}
+
+{{#if (eq deployment.backend.platform "heroku")}}
+### Heroku
+
+```bash
+# Push to Heroku
+git push heroku {{versionControl.defaultBranch}}
+
+# Check logs
+heroku logs --tail
+```
+{{/if}}
+
+{{#if (eq deployment.backend.platform "render")}}
+### Render
+
+```bash
+# Render auto-deploys from git push
+git push origin {{versionControl.defaultBranch}}
+
+# Or trigger manual deploy via API
+curl -X POST "https://api.render.com/v1/services/{{deployment.backend.serviceId}}/deploys" \
+  -H "Authorization: Bearer $RENDER_API_KEY"
+```
+{{/if}}
+
+{{#if (eq deployment.backend.platform "fly")}}
+### Fly.io
+
+```bash
+# Deploy to Fly
+fly deploy
+
+# Check status
+fly status
+```
+{{/if}}
+
+{{#if (eq deployment.backend.platform "vercel")}}
+### Vercel (Serverless)
+
+```bash
+vercel --prod
+```
+{{/if}}
+
+{{#if (eq deployment.backend.platform "self-hosted")}}
+### Self-Hosted
+
+```bash
+# SSH deploy
+ssh {{deployment.backend.selfHostedConfig.sshUser}}@{{deployment.backend.selfHostedConfig.sshHost}} -p {{deployment.backend.selfHostedConfig.sshPort}} \
+  'cd {{deployment.backend.selfHostedConfig.appPath}} && git pull && {{deployment.backend.selfHostedConfig.deployScript}}'
+```
+{{/if}}
+
+---
+
+## Frontend Deployment
+
+{{#if (eq deployment.frontend.platform "cloudflare")}}
+### Cloudflare Pages
+
+```bash
+# Clean and build
+rm -rf {{frontend.distDir}}
+{{frontend.buildCommand}}
+
+# Deploy
+npx wrangler pages deploy {{frontend.distDir}} --project-name={{deployment.frontend.projectName}}
+```
+{{/if}}
+
+{{#if (eq deployment.frontend.platform "vercel")}}
+### Vercel
+
+```bash
+vercel --prod
+```
+{{/if}}
+
+{{#if (eq deployment.frontend.platform "netlify")}}
+### Netlify
+
+```bash
+# Build
+{{frontend.buildCommand}}
+
+# Deploy
+netlify deploy --prod --dir={{frontend.distDir}}
+```
+{{/if}}
+
+{{#if (eq deployment.frontend.platform "github-pages")}}
+### GitHub Pages
+
+```bash
+# Build and deploy (assumes gh-pages package)
+{{frontend.buildCommand}}
+npm run deploy
+```
+{{/if}}
+
+{{#if (eq deployment.frontend.platform "railway")}}
+### Railway (Static)
+
+```bash
+# Push triggers auto-deploy
+git push origin {{versionControl.defaultBranch}}
+```
+{{/if}}
+
+---
+
+## Verification
+
+### Backend Health Check
+
+```bash
+curl {{deployment.backend.productionUrl}}{{backend.healthEndpoint}}
+```
+
+### Frontend Verification
+
+1. Open {{deployment.frontend.productionUrl}}
+2. Verify page loads correctly
+3. Check browser console for errors
+4. Test critical user flows
+
+### E2E Smoke Test (Optional)
+
+```bash
+{{#if testing.e2e.testCommand}}
+{{testing.e2e.testCommand}} --grep "smoke"
+{{else}}
+npx playwright test --grep "smoke"
+{{/if}}
+```
+
+---
+
+## Rollback Procedures
+
+### Backend Rollback
+
+{{#if (eq deployment.backend.platform "railway")}}
+```javascript
+// List recent deployments
+mcp__railway-mcp-server__deployment_list({
+  projectId: "{{deployment.backend.projectId}}",
+  serviceId: "{{deployment.backend.serviceId}}"
+})
+
+// Redeploy previous version
+mcp__railway-mcp-server__deployment_rollback({
+  projectId: "{{deployment.backend.projectId}}",
+  serviceId: "{{deployment.backend.serviceId}}",
+  deploymentId: "<previous-deployment-id>"
+})
+```
+{{else}}
+```bash
+# Revert commit and redeploy
+git revert HEAD
+git push origin {{versionControl.defaultBranch}}
+```
+{{/if}}
+
+### Frontend Rollback
+
+```bash
+# Revert to previous build
+git revert HEAD
+{{frontend.buildCommand}}
+{{deployment.frontend.deployCommand}}
+```
+
+{{else}}
+
+## Frontend Deployment Not Configured
+
+Frontend deployment platform is not configured.
+
+Run `/menu` → Project Settings → Deployment Platforms to configure.
+
+{{/if}}
+{{else}}
+
+## Backend Deployment Not Configured
+
+Backend deployment platform is not configured.
+
+Run `/menu` → Project Settings → Deployment Platforms to configure.
+
+{{/if}}
+
+---
+
+*Generated from tech-stack.json template*