# WDMaker Execution Decision Trees **Purpose**: Visual decision logic for all critical execution points **Audience**: Operators, decision-makers, troubleshooters **Status**: Ready for autonomous execution with human oversight --- ## Decision Tree 1: "Should I Execute Batch 001 Finalization Now?" ``` START: Current time, you want to finalize batch 001 │ ├─ QUESTION: Have I confirmed I-status = 517? │ ├─ NO → WAIT │ │ Action: Check again in 30 minutes │ │ Command: tools/shared/list-sites.sh --batch 001 --status "I" | wc -l │ │ Next: Come back to this tree │ │ │ └─ YES → PROCEED │ Continue to next question │ ├─ QUESTION: Do I understand finalization is idempotent? │ ├─ NO → READ │ │ Document: FINALIZATION_EXECUTION_GUIDE.md (section: Idempotency) │ │ Time: 5 minutes │ │ Then return to this tree │ │ │ └─ YES → PROCEED │ Continue to next question │ ├─ QUESTION: Have I reviewed the pre-finalization checks? │ ├─ NO → READ │ │ Document: FINALIZATION_EXECUTION_GUIDE.md (section: Pre-Finalization) │ │ Time: 10 minutes │ │ Then return to this tree │ │ │ └─ YES → PROCEED │ Continue to next question │ ├─ QUESTION: Is the system stable (no obvious problems)? │ ├─ NO → DIAGNOSE │ │ Use: COMPREHENSIVE_TROUBLESHOOTING_MATRIX.md │ │ Or: EMERGENCY_RESPONSE_GUIDE.md │ │ Time: 15-30 minutes │ │ Then return to this tree │ │ │ └─ YES → READY TO EXECUTE │ Go to Section: Finalization Execution Steps │ Expected Duration: 5 minutes │ Expected Outcome: All 517 sites at Q status RESULT: Execute finalization → Proceed to Batch 010 ``` --- ## Decision Tree 2: "Batch 001 Finalization Failed - What Now?" ``` START: finish.sh command returned error │ ├─ QUESTION: What type of error occurred? │ │ │ ├─ ERROR TYPE: "Registry not found" or "Permission denied" │ │ Action: │ │ 1. Check: ls -l .smbatcher/REGISTRY.md │ │ 2. Check: ls -l .smbatcher/batches/Batch_001.md │ │ 3. If missing: Recover from git history │ │ git checkout .smbatcher/REGISTRY.md │ │ 4. Retry finalization │ │ Resolution: Low risk, safe to retry │ │ │ ├─ ERROR TYPE: "Sites with status != I" or "Transition failed" │ │ Meaning: Some sites not ready for finalization │ │ Action: │ │ 1. Check: tools/shared/list-sites.sh --batch 001 | grep -v "I" │ │ 2. Identify status distribution │ │ 3. If only a few sites != I: │ │ - Wait 5 more minutes for autonomous completion │ │ - Recheck I-status │ │ - Retry finalization │ │ 4. If many sites != I: │ │ - Escalate to EMERGENCY_RESPONSE_GUIDE.md │ │ - Check system resource status │ │ Resolution: Medium risk, may need wave redeployment │ │ │ ├─ ERROR TYPE: "Cannot write to registry" or "Atomic operation failed" │ │ Meaning: Concurrent access issue or disk problem │ │ Action: │ │ 1. Check: df -h .smbatcher/ │ │ 2. Check: ls -la .smbatcher/ (for lock files) │ │ 3. If lock file exists: │ │ - Wait 30 seconds │ │ - Remove: rm -f .smbatcher/.lock │ │ - Retry finalization │ │ 4. If disk space low: │ │ - Clear temporary files │ │ - Check: du -sh .smbatcher/tmp/ │ │ - Then retry │ │ Resolution: Medium risk, usually recoverable │ │ │ └─ ERROR TYPE: Unknown or system error │ Action: │ 1. Consult EMERGENCY_RESPONSE_GUIDE.md (Scenario 3) │ 2. Run: tools/check/status-report.sh │ 3. Document error message completely │ 4. Attempt retry (finalization is idempotent) │ Resolution: Follow guide, escalate if needed │ ├─ RETRY DECISION: Should I retry finalization? │ ├─ YES, error seems transient → │ │ Action: tools/implement/finish.sh --batch 001 --root . │ │ Expected: Success on retry │ │ If fails again: Move to ESCALATION │ │ │ └─ NO, error seems systemic → │ ESCALATION REQUIRED │ Action: Read EMERGENCY_RESPONSE_GUIDE.md completely │ Then: Implement recommended recovery procedure RESULT: Either finalization succeeds OR escalation path identified ``` --- ## Decision Tree 3: "Ready to Start Batch 010? Check These First" ``` START: Batch 001 successfully finalized (all 517 at Q status) │ ├─ CONFIRM: Batch 001 finalization is complete │ Check: tools/shared/list-sites.sh --batch 001 --status "Q" | wc -l │ Expected: 517 │ ├─ NOT 517 → ERROR │ │ Something went wrong with finalization │ │ Go back to Decision Tree 2 (Finalization Failed) │ │ │ └─ IS 517 → PROCEED │ ├─ CONFIRM: Batch 010 unassigned site exists │ Check: tools/shared/list-sites.sh --status "-" | grep "20241204" │ Expected: 1 site (20241204.com) │ ├─ NOT FOUND → ERROR │ │ Site may already be assigned │ │ Check: tools/shared/list-sites.sh | grep "20241204" │ │ If already processed: Skip batch 010, go to completion │ │ │ └─ FOUND → PROCEED │ ├─ ESTIMATE: Time available for batch 010 │ Expected duration: 30-60 minutes │ Question: Do you have time now? │ ├─ YES → PROCEED to batch 010 execution │ │ Follow: BATCH_010_DETAILED_WORKFLOW.md │ │ │ └─ NO → SCHEDULE for later │ Save this decision tree state │ Come back when ready │ ├─ CONFIRM: System is stable and ready │ Check: tools/check/status-report.sh │ Expected: No errors, disk space available (>1GB), memory available (>4GB) │ ├─ PROBLEMS FOUND → DIAGNOSE │ │ Use: COMPREHENSIVE_TROUBLESHOOTING_MATRIX.md │ │ Fix issues first │ │ Then come back to this tree │ │ │ └─ STABLE → READY TO PROCEED │ Follow: BATCH_010_DETAILED_WORKFLOW.md │ Expected: 20241204.com processed in 30-60 minutes │ Expected outcome: Site at Q status RESULT: Either Batch 010 execution ready OR system issues identified ``` --- ## Decision Tree 4: "How to Handle Stuck Sites During Batch 001" ``` START: Some sites appear stuck (status not changing for 45+ minutes) │ ├─ DIAGNOSIS: Identify which sites are stuck │ Check: tools/shared/list-sites.sh --batch 001 --status "i" │ Question: Are any sites at i (in-progress)? │ ├─ NO (all at O or I) → Progress is normal │ │ Continue monitoring │ │ Check again in 30 minutes │ │ │ └─ YES (sites at i) → Potential issue │ Continue to next step │ ├─ INVESTIGATION: Are agents still running? │ Check: ps aux | grep -i opus (or check agent dashboard) │ ├─ NO AGENTS RUNNING → PROBLEM │ │ Waves may have terminated prematurely │ │ Action: Consult EMERGENCY_RESPONSE_GUIDE.md (Scenario 1) │ │ Possible solution: Redeploy waves 4-9 │ │ │ └─ AGENTS RUNNING → Continue investigation │ ├─ CHECK: Are these specific sites or all sites? │ ├─ SPECIFIC SITES (< 5 stuck) │ │ Action: │ │ 1. Check site-specific logs (if available) │ │ 2. See if files were generated despite status not updating │ │ 3. Manually update status if files exist: │ │ - Mark as I status │ │ - Registry update via complete.sh │ │ │ └─ ALL/MANY SITES (> 10 stuck) │ Action: │ 1. Consult COMPREHENSIVE_TROUBLESHOOTING_MATRIX.md │ 2. Check system resources (CPU, memory, disk) │ 3. Possible issue: Resource exhaustion │ 4. Solution: Wait for resource recovery OR restart agents │ ├─ DECISION: Wait or intervene? │ ├─ I'LL WAIT │ │ Reason: Agents still active, might complete soon │ │ Action: Check again in 30 minutes │ │ Condition: Only if < 5% of batch stuck │ │ │ └─ I'LL INTERVENE │ Reason: Too many sites stuck, system problem suspected │ Action: Follow EMERGENCY_RESPONSE_GUIDE.md recovery procedure │ Condition: If > 10% stuck OR > 45 minutes with no progress RESULT: Either wait for autonomous completion OR invoke recovery ``` --- ## Decision Tree 5: "Project Completion - 99.6% Verification" ``` START: Both Batch 001 and Batch 010 finalized │ ├─ COUNT: Verify total finalized sites │ Check: tools/shared/list-sites.sh --status "Q" | wc -l │ Expected range: 566-568 │ │ │ ├─ COUNT = 568 → PERFECT │ │ All sites processed │ │ Completion: 100% of catalog │ │ Action: Document as EXCELLENT outcome │ │ │ ├─ COUNT = 566 → EXPECTED │ │ 99.6% completion (standard target) │ │ Missing: 2 sites │ │ Reason: Possible batch 001 failures in waves 1-3 │ │ Action: Identify missing 2 sites, document why │ │ Command: tools/shared/list-sites.sh | grep -v "^| .*| Q |" │ │ │ └─ COUNT < 566 → UNEXPECTED │ Possible issues with finalization │ Action: Run complete status audit │ Command: tools/shared/list-sites.sh │ Then consult EMERGENCY_RESPONSE_GUIDE.md for recovery │ ├─ IDENTIFY: Missing sites (if count < 568) │ Check: tools/shared/list-sites.sh | grep -E "^\| [^|]* \| [^Q]" │ Action: List domain names and their current status │ Categorize: Are they stuck at O, i, or I? │ │ │ ├─ STUCK AT O (Open) │ │ Meaning: Never entered implementation │ │ Reason: Likely didn't get wave assignment │ │ Recovery: May need manual requeue OR accept as limitation │ │ │ ├─ STUCK AT i (In-progress) │ │ Meaning: Agent started but didn't complete │ │ Reason: Agent timeout or error │ │ Recovery: Could retry agents on these sites │ │ │ └─ STUCK AT I (Implemented) │ Meaning: Files generated but finalization didn't mark Q │ Reason: Finalization issue on these specific sites │ Recovery: Attempt manual finalization for these sites │ ├─ DECISION: Is 99.6% acceptable? │ ├─ YES → ACCEPT AND COMPLETE │ │ Action: Mark project as COMPLETE │ │ Note: Document which 2 sites missed and why │ │ Review: LESSONS_LEARNED_AND_RECOMMENDATIONS.md │ │ │ └─ NO → ATTEMPT RECOVERY │ Action: Follow recovery steps above │ Risk: May not recover missing sites │ Time: 30+ minutes RESULT: Project completion documented, status finalized ``` --- ## Decision Tree 6: "System Resource Emergency - What to Do?" ``` START: System running low on resources (disk, memory, or CPU) │ ├─ SEVERITY: How bad is it? │ │ │ ├─ DISK: Free space < 500MB │ │ ├─ SEVERE (< 200MB free) │ │ │ Risk: System might crash, registry corruption possible │ │ │ Action: │ │ │ 1. IMMEDIATELY stop new agents: Don't deploy more waves │ │ │ 2. Clear temporary files: rm -rf .smbatcher/tmp/* │ │ │ 3. Check: du -sh sites/*/ │ │ │ 4. Consider: Archiving completed batches to save space │ │ │ 5. If critical: Pause operations until disk space recovered │ │ │ │ │ └─ MODERATE (200-500MB free) │ │ Risk: Some operations may fail │ │ Action: │ │ 1. Monitor disk space closely │ │ 2. Don't archive or copy large files │ │ 3. Complete current work ASAP │ │ 4. Plan cleanup for later │ │ │ ├─ MEMORY: Free memory < 1GB │ │ ├─ SEVERE (< 500MB free) │ │ │ Risk: Agents may timeout, system becomes slow │ │ │ Action: │ │ │ 1. Reduce agents per wave: Don't deploy new waves yet │ │ │ 2. Wait for running agents to complete │ │ │ 3. Restart system if needed (last resort) │ │ │ 4. Current work: Allow to complete, then pause │ │ │ │ │ └─ MODERATE (500MB-1GB free) │ │ Risk: Slower than expected │ │ Action: │ │ 1. Monitor memory usage │ │ 2. Reduce new agent deployments slightly │ │ 3. Complete current work normally │ │ │ └─ CPU: Usage > 90% sustained │ ├─ SEVERE (98-100% consistent) │ │ Risk: System may become unresponsive │ │ Action: │ │ 1. Stop non-critical processes (if safe) │ │ 2. Agents will still work but slower │ │ 3. Expected: Return to normal after current wave completes │ │ │ └─ MODERATE (70-90%) │ Risk: System overloaded but functional │ Action: │ 1. This is normal during wave execution │ 2. No action needed, expected behavior │ ├─ DECISION: Can current work continue? │ ├─ YES → Monitor but continue │ │ Check status every 15-30 minutes │ │ Watch for further resource degradation │ │ │ └─ NO → Pause and Recover │ Action: │ 1. Don't deploy new waves │ 2. Let current agents finish │ 3. Identify root cause (large files? leak?) │ 4. Fix before continuing │ ├─ ESCALATION: Contact system admin if: │ ✓ Disk space < 200MB and can't be freed │ ✓ Memory consistently < 200MB │ ✓ System becoming unresponsive │ Document: Current resource usage, recent changes RESULT: Recovery actions taken, operations adjusted ``` --- ## Quick Reference: Decision Tree Selection Guide **"What should I do right now?"** → Use this chart: | Current Situation | Decision Tree | Time Needed | |------------------|---------------|-------------| | Want to finalize batch 001 | Tree 1 | 15-30 min | | Finalization failed | Tree 2 | 30-60 min | | Ready to start batch 010 | Tree 3 | 10 min | | Sites appear stuck | Tree 4 | 30 min | | Project completed | Tree 5 | 20 min | | System running low on resources | Tree 6 | 20-30 min | --- ## Using These Decision Trees 1. **Find your situation** in the Quick Reference table 2. **Follow the tree** from START to RESULT 3. **Answer questions honestly** at each branch point 4. **Take recommended actions** immediately 5. **Record outcomes** for lessons learned Each tree is designed to reach a conclusion within 15-60 minutes. If you need to escalate or get more information, the trees reference specific guides (EMERGENCY_RESPONSE_GUIDE.md, etc.). --- *Decision Trees: 2026-03-24* *Purpose: Real-time decision support for all critical execution points* *Confidence: Based on proven system behavior across 568-site catalog*