CRITICAL: Windows reboot protection commit Files committed: - STACKCP_S2_SWARM_DEPLOYMENT.md (599 lines) - USER-TESTING-GUIDE.md - WAKE_UP_SUMMARY.md (504 lines) - sample-agent-tasks-tier1.json - test-results-feature-selector.md Current status: - User selected ALL 11 features with detailed notes - 4-hour MVP strategy: Photo inventory + Document search + Maintenance timeline - GitHub as main codebase location - Need agent communication setup (GitHub Issues + AUTONOMOUS-NEXT-TASKS.md) Next: Expand feature selector with full research (94 files) + 5 session prompts IF.TTT Citation: if://checkpoint/pre-reboot-2025-11-13 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
6.5 KiB
6.5 KiB
Feature Selector User Testing Guide
Quick Start: https://digital-lab.ca/navidocs/builder/
Quick Test (5 minutes)
Step 1: Open Feature Selector
Open browser: https://digital-lab.ca/navidocs/builder/
Step 2: Select Tier 1 Features
Check all 4 Tier 1 features (red badges):
- Photo-Based Inventory Tracking
- Smart Maintenance Tracking & Reminders
- Document Versioning & Audit Trail
- Multi-User Expense Tracking
Step 3: Export Agent Tasks
- Click green "Export Agent Tasks" button
- Verify alert shows:
- Total tasks: 21
- P0 tasks: 21
- Estimated hours: 53h
- Download:
navidocs-agent-tasks-2025-11-13.json
Step 4: Verify JSON Structure
Open downloaded file and check:
metadata.selected_features: 4agents.agent-1-backend.tasks: 4 tasksagents.agent-2-frontend.tasks: 4 tasksagents.agent-3-database.tasks: 4 tasksagents.agent-4-integration.tasks: 1 task (expense-tracking only)agents.agent-5-testing.tasks: 4 taskssummary.total_tasks: 21summary.estimated_total_hours: 53
Expected Result: ✅ JSON matches sample in /home/setup/navidocs/sample-agent-tasks-tier1.json
Full Test (15 minutes)
Test Case 1: No Features Selected
- Clear all selections (Click "Clear All")
- Click "Export Agent Tasks"
- Expected: Alert "Please select at least one feature before exporting agent tasks!"
Test Case 2: Single Feature
- Select only "Photo-Based Inventory Tracking" (Tier 1)
- Click "Export Agent Tasks"
- Expected: 5 tasks (1 per agent, except agent-4 skips this feature)
Test Case 3: Mix of Tiers
- Select:
- Tier 1: Photo-Based Inventory Tracking (Must-Have: 10)
- Tier 2: Camera Integration (Must-Have: 7)
- Tier 3: WhatsApp Integration (Must-Have: 5)
- Click "Export Agent Tasks"
- Expected:
- P0 tasks: 5 (from Tier 1, Must-Have 10 → P0)
- P1 tasks: 5 (from Tier 2, Must-Have 7 → P1)
- P2 tasks: 5 (from Tier 3, Must-Have 5 → P2)
Test Case 4: Priority Calculation
- Select "Camera Integration" (Tier 2, default Must-Have: 7)
- Adjust slider to 9
- Click "Export Agent Tasks"
- Expected: All 5 tasks now P0 (Must-Have 9 ≥ 8 → P0)
Test Case 5: Integration Features
- Select all features requiring third-party integrations:
- Camera Integration (Home Assistant)
- WhatsApp Integration (WhatsApp Business API)
- Accounting Integration (Spliit fork)
- Multi-User Expense Tracking (OCR + bank API)
- Click "Export Agent Tasks"
- Expected: agent-4-integration has 4 tasks (one for each integration feature)
Test Case 6: All Features
- Click "Select All" (11 features)
- Click "Export Agent Tasks"
- Expected:
- Total tasks: 54-55 (11 features × ~5 agents, minus skipped integrations)
- P0 tasks: ~35-40 (features with Must-Have ≥ 8)
- Estimated hours: ~150-200h
Validation Checklist
Feature Display
- 11 features rendered correctly
- Tier badges visible (red=Tier 1, orange=Tier 2, green=Tier 3)
- Must-Have rating displayed (X/10)
- Priority badges (CRITICAL/HIGH/MEDIUM)
- Savings amounts shown
Interaction
- Checkboxes toggle feature selection
- Selected features highlight in blue (#F0F8FF)
- Stats update in real-time (Selected count, Avg Must-Have, Total Savings)
- Slider adjusts Must-Have rating (1-10)
- Notes textarea accepts input
Buttons
- "Select All" checks all 11 features
- "Clear All" unchecks all features
- "Export JSON" downloads feature-selection JSON
- "Export Agent Tasks" downloads agent-tasks JSON (green button)
- "Print" opens print dialog
JSON Structure (Agent Tasks)
{
"metadata": {
"timestamp": "ISO 8601",
"meeting": "Riviera Plaisance Partnership",
"deployment_target": "StackCP shared hosting",
"selected_features": <count>,
"swarm_pattern": "S2 (5 Haiku agents parallel)",
"autonomous_task_file": "AUTONOMOUS-NEXT-TASKS.md"
},
"agents": {
"agent-1-backend": { "role": "...", "model": "haiku", "tasks": [...] },
"agent-2-frontend": { "role": "...", "model": "haiku", "tasks": [...] },
"agent-3-database": { "role": "...", "model": "haiku", "tasks": [...] },
"agent-4-integration": { "role": "...", "model": "haiku", "tasks": [...] },
"agent-5-testing": { "role": "...", "model": "haiku", "tasks": [...] }
},
"summary": {
"total_tasks": <count>,
"estimated_total_hours": <hours>,
"p0_tasks": <count>,
"deployment_instructions": [...]
}
}
Task Structure
Each task must have:
feature_id: String (e.g., "inventory-tracking")title: String (e.g., "API endpoints for Photo-Based Inventory Tracking")priority: String (P0/P1/P2)status: String ("pending")description: String (technical description)technical_notes: String (from feature.description)user_notes: String (from notes textarea or "None")estimated_hours: Number (2-4 hours depending on tier)dependencies: Array of strings (e.g., ["database schema ready"])
Known Issues
❌ Issue 1: WebFetch Network Access
- Description: Agent 3 WebFetch tool cannot access live deployment
- Impact: Cannot verify page loads in browser (static analysis only)
- Workaround: Manual browser testing required
- Status: P2 (Medium) - File analysis confirms correctness
Success Criteria
Minimum Viable Test (MVP)
- Page loads at https://digital-lab.ca/navidocs/builder/
- Can select/deselect features
- "Export Agent Tasks" button works
- JSON downloads with correct structure
- Alert shows task summary
Production Ready
- All 6 test cases pass
- JSON validates against schema
- No console errors
- Works on mobile (responsive)
- Print view formats correctly
Next Steps After Testing
If Tests Pass
- Deploy agent-tasks JSON to StackCP
- Launch 5 parallel Haiku agents
- Agents poll AUTONOMOUS-NEXT-TASKS.md
- Begin implementation (Backend → Database → Frontend → Integration → Testing)
If Tests Fail
- Document failures in
/home/setup/navidocs/test-failures.md - Report to Agent 3 for debugging
- Fix bugs in feature-selector.html (lines 591-759)
- Redeploy to StackCP
- Retest
Sample Output
See /home/setup/navidocs/sample-agent-tasks-tier1.json for expected JSON structure when all 4 Tier 1 features are selected.
Generated by: Agent 3 - Feature Selector Testing & Validation (Haiku) Date: 2025-11-13 IF.TTT Compliance: Testing instructions traceable to feature-selector.html:591-759