[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6CQYHV1W1X6J09WM4B35YP7":3,"skill-same:01M6CQYHV1W1X6J09WM4B35YP7":73},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":29,"linkOnly":30,"markdownContent":31,"markdownBody":32,"frontMatter":33,"refName":34,"commitSha":35,"skillDirectory":36,"forksCount":14,"repositoryHtmlUrl":37,"repositoryHomepageUrl":6,"ownerHtmlUrl":38,"ownerBlogUrl":39,"ownerDescription":6,"updatedAt":40,"install":41,"resources":54},"01M6CQYHV1W1X6J09WM4B35YP7","task-processor-parallel",null,"Process tasks using async background subagents for massive parallelization and speedup.","dundas\u002Fagentbootup-public","dundas","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F627426?v=4","01M6CQVP2KHXEP7JH9WKKF59S7","David K. Dundas","01M6CQVP76D48NCMMXYDK9CBA8",0,"MIT","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Ftask-processor-parallel\u002FSKILL.md",[],[19,20,21,22,23,24,25,26,27,28],"processor","parallel","task","task processor parallel","workflow","tasks","parallelization","process","subagents","async","2026-11-10 10:12:30",false,"---\nname: task-processor-parallel\ndescription: Process tasks using async background subagents for massive parallelization and speedup.\ncategory: workflow\n---\n\n# Parallel Task Processor with Async Subagents\n\n## Prerequisites\n\n**Required:**\n- Claude Code v2.0.60+ (for async subagent support)\n- Git repository with remote configured\n- GitHub CLI (`gh`) installed and authenticated: `gh auth login`\n- GitHub repository (PR automation features require GitHub)\n\n**Optional:**\n- CI\u002FCD configured for automated checks\n- Sufficient API quota for parallel subagents\n\n**Non-GitHub Hosting:** If using GitLab, Bitbucket, or other platforms, the PR automation features won't work. You can still use parallel task processing, but create PRs manually.\n\n## Overview\nLeverages Claude Code's async subagents (v2.0.60+) to process multiple tasks in parallel, dramatically reducing implementation time.\n\n## Key Features\n- **Parallel Execution**: Spawn multiple subagents to work on independent tasks simultaneously\n- **Background Mode**: Subagents run in background, don't block orchestrator\n- **Auto-Notification**: Subagents notify orchestrator when complete\n- **PR Automation**: Automated gap analysis and PR comments after each phase\n\n## Parallelization Strategy\n\n### Level 1: Phase-Level Parallelization\nLaunch multiple phases in parallel when no dependencies exist. Use Claude Code's Task tool with multiple parallel calls:\n\n```\n\u002F\u002F Batch 1: Independent phases (run in parallel)\n\u002F\u002F Send a single message with multiple Task tool calls:\n\nTask(subagent_type: \"tdd-developer\", description: \"Phase 1: Comparison\", prompt: \"...\")\nTask(subagent_type: \"tdd-developer\", description: \"Phase 5: Performance\", prompt: \"...\")\nTask(subagent_type: \"tdd-developer\", description: \"Phase 8: Notifications\", prompt: \"...\")\n\n\u002F\u002F Wait for all to complete, then launch dependent phases\n```\n\n### Level 2: Sub-Task Parallelization\nWithin a single phase, parallelize independent sub-tasks:\n\n```\n\u002F\u002F Phase 1 split into 3 parallel streams (single message, multiple Task calls):\n\nTask(subagent_type: \"tdd-developer\", description: \"Discovery + Normalization\", prompt: \"Tasks 1.1, 1.2...\")\nTask(subagent_type: \"tdd-developer\", description: \"Ranking + Reasoning\", prompt: \"Tasks 1.3, 1.4...\")\nTask(subagent_type: \"tdd-developer\", description: \"Types + Tests\", prompt: \"Tasks 1.5, 1.6...\")\n\n\u002F\u002F When all complete → Final task\nTask(subagent_type: \"tdd-developer\", description: \"Create PR\", prompt: \"Task 1.7...\")\n```\n\n## Execution Protocol\n\n### 1. Analyze Dependencies\n```markdown\nPhase Dependencies:\n- Phase 1: None → Start immediately\n- Phase 2: Depends on Phase 1\n- Phase 3: Depends on Phase 2\n- Phase 4: Depends on Phase 3\n- Phase 5: None → Start immediately (parallel to 1-3)\n- Phase 6: Depends on Phase 5\n- Phase 7: Depends on Phase 3 + Phase 5\n- Phase 8: None → Start immediately (parallel to all)\n- Phase 9: Depends on all phases\n```\n\n### 2. Create Execution Batches\n```markdown\nBatch 1 (Parallel):\n- Phase 1 (large effort)\n- Phase 5 (medium effort)\n- Phase 8 (small effort)\n→ Total: Large effort (instead of 3x sequential!)\n\nBatch 2 (Parallel - after Phase 1):\n- Phase 2 (large effort)\n- Phase 6 (small effort, after Phase 5)\n→ Total: Large effort\n\nBatch 3 (Parallel - after Phase 2):\n- Phase 3 (large effort)\n→ Total: Large effort\n\nBatch 4 (Parallel - after Phase 3):\n- Phase 4 (medium effort)\n- Phase 7 (large effort, also needs Phase 5)\n→ Total: Large effort\n\nBatch 5 (Sequential - after all):\n- Phase 9 (medium effort)\n→ Total: Medium effort\n\n**Parallel execution reduces total effort by ~25% compared to sequential**\n```\n\n### 3. Spawn Async Subagents\n\n**Using Task Tool (parallel calls in single message):**\n\nTo launch multiple phases in parallel, send a single message with multiple Task tool calls:\n\n```\nTask tool call 1:\n  subagent_type: \"tdd-developer\"\n  description: \"Implement Phase 1: Comparison\"\n  prompt: \"Work through tasks 1.1 through 1.7 from tasks\u002F[task-file].md\n\n    Tasks:\n    - 1.1: Extend discovery engine\n    - 1.2: Create score normalization\n    - 1.3: Build ranking algorithm\n    - 1.4: Add comparison reasoning\n    - 1.5: Create TypeScript types\n    - 1.6: Integration tests\n    - 1.7: Create PR\n\n    Follow test-driven development.\n    Commit after each task.\n    When complete, create PR and run gap analysis.\"\n\nTask tool call 2 (in same message for parallel execution):\n  subagent_type: \"tdd-developer\"\n  description: \"Implement Phase 5: Performance\"\n  prompt: \"Work through tasks 5.1 through 5.7...\"\n\n\u002F\u002F Both subagents run in parallel\n\u002F\u002F Results returned when each completes\n```\n\n### 4. Monitor Progress\n\n**Check Subagent Status:**\n```bash\n# List running background tasks\n\u002Ftasks\n\n# Background tasks complete and return results automatically\n# Monitor the conversation for task completion notifications\n```\n\n### 5. Collect Results\n\nWhen a subagent completes, it notifies the orchestrator:\n```typescript\n\u002F\u002F Subagent 1 completes Phase 1\nResult: {\n  phase: 1,\n  pr_number: 16,\n  files_changed: 6,\n  tests_added: 50,\n  gap_analysis: \"docs\u002FPR_16_GAP_ANALYSIS.md\",\n  status: \"ready_for_review\"\n}\n\n\u002F\u002F Orchestrator receives notification\n\u002F\u002F Triggers next dependent phase (Phase 2)\n```\n\n### 6. Automated PR Workflow (Per Subagent)\n\nEach subagent follows this protocol when its phase completes:\n\n1. **Run Tests**: Full test suite for the phase\n2. **Create Branch**: `git checkout -b feat\u002Fphase-X-name`\n3. **Push**: `git push -u origin feat\u002Fphase-X-name`\n4. **Create PR**: Using `gh pr create`\n5. **Wait for CI**: 30-second delay\n6. **Check Status**: `gh pr view` + `gh pr checks`\n7. **Generate Gap Analysis**: Create `docs\u002FPR_X_GAP_ANALYSIS.md`\n8. **Push Gap Analysis**: Commit and push\n9. **Add PR Comment**: Detailed changes summary\n10. **Notify Orchestrator**: Return results\n\n## Example: Full Autonomous Run\n\n```bash\n# User starts the processor\nskill: task-processor-parallel\n\n# Orchestrator analyzes dependencies\n→ Identifies 5 execution batches\n→ Creates parallelization plan\n\n# Batch 1: Launch 3 phases in parallel\n→ Spawn Phase 1 subagent (tdd-developer)\n→ Spawn Phase 5 subagent (tdd-developer)\n→ Spawn Phase 8 subagent (tdd-developer)\n\n# Orchestrator: \"3 subagents launched, monitoring progress...\"\n\n# [After Phase 5 completes first - shortest effort]\n# Phase 5 completes first (shortest)\n→ Subagent 5 creates PR #20\n→ Subagent 5 runs gap analysis\n→ Subagent 5 notifies orchestrator: \"Phase 5 done, PR #20 ready\"\n\n# Phase 1 completes\n→ Subagent 1 creates PR #16\n→ Subagent 1 runs gap analysis\n→ Subagent 1 notifies orchestrator: \"Phase 1 done, PR #16 ready\"\n→ Orchestrator triggers Phase 2 (depends on Phase 1)\n\n# Phase 8 completes\n→ Subagent 8 creates PR #23\n→ Subagent 8 runs gap analysis\n→ Subagent 8 notifies orchestrator: \"Phase 8 done, PR #23 ready\"\n\n# Batch 2: Launch Phase 2 (depends on Phase 1 - just completed)\n→ Spawn Phase 2 subagent (tdd-developer)\n\n# ... Continue through all batches ...\n\n# All phases complete\n→ Orchestrator: \"All 9 PRs created and reviewed!\"\n→ User reviews PRs and merges\n```\n\n## Benefits\n\n### Time Savings\n- **Sequential**: Sum of all phase efforts\n- **Parallel**: Longest path through dependency graph\n- **Savings**: ~25% reduction in total effort\n\n### Efficiency\n- No idle time waiting for sequential tasks\n- Better resource utilization\n- Continuous progress across multiple fronts\n\n### Flexibility\n- Can pause\u002Fresume individual subagents\n- Can prioritize critical phases\n- Can adjust parallelization on the fly\n\n## Limitations & Considerations\n\n### Max Concurrent Subagents\nRecommended limits to avoid overwhelming resources:\n\n| Scenario | Max Parallel | Reason |\n|----------|--------------|--------|\n| Small tasks (\u003C 30 min each) | 4-5 | Token budget, context quality |\n| Medium tasks (1-2 hours) | 2-3 | Avoid merge conflicts |\n| Large tasks (half day+) | 1-2 | Resource contention |\n\n**Guidelines:**\n- Start conservative (2-3 parallel) and increase if stable\n- Monitor token usage with `\u002Fstats` command\n- If subagents produce poor results, reduce parallelism\n- Complex codebases benefit from less parallelism\n\n### Context Cost\n- Each subagent has its own context window\n- More parallel agents = higher token usage\n- Monitor costs with `\u002Fstats`\n\n### Merge Conflicts\n- Parallel work may touch same files\n- Mitigate: Design tasks to minimize overlap\n- Use clear file ownership per phase\n\n### CI Load\n- Multiple PRs may overload CI\n- Mitigate: Stagger PR creation\n- Use draft PRs to delay CI\n\n## Troubleshooting Parallel Subagents\n\n### Subagent Fails or Produces Poor Results\n\n**Symptoms:**\n- Subagent returns incomplete work\n- Code doesn't compile or tests fail\n- Subagent seems \"lost\" or unfocused\n\n**Diagnosis:**\n1. Check task complexity - may be too large for single subagent\n2. Check task dependencies - may need output from another phase\n3. Check context - subagent may lack necessary codebase knowledge\n\n**Solutions:**\n- Break task into smaller sub-tasks\n- Run the phase sequentially instead of parallel\n- Provide more context in the subagent prompt\n- Reduce number of parallel subagents\n\n### Subagent Hangs or Times Out\n\n**Symptoms:**\n- `\u002Ftasks` shows subagent running for unusually long time\n- No progress updates\n\n**Solutions:**\n1. Check `\u002Ftasks` for status\n2. If stuck, cancel and retry with smaller scope\n3. Consider running that phase sequentially\n\n### Merge Conflicts Between Parallel Phases\n\n**Symptoms:**\n- `git push` fails with conflict errors\n- PRs can't be merged\n\n**Solutions:**\n1. Identify which files conflict\n2. Determine which phase's changes are correct\n3. Manually resolve conflicts\n4. Re-run validation on merged code\n5. **Prevention:** Assign clear file ownership per phase\n\n### Inconsistent State Across Phases\n\n**Symptoms:**\n- Phase A created interface X\n- Phase B expected different interface\n- Integration fails\n\n**Solutions:**\n1. Define interfaces\u002Fcontracts BEFORE parallel execution\n2. Create shared types\u002Finterfaces in a prerequisite phase\n3. Re-run dependent phase after fixing interface\n\n### Debugging Checklist\n\nWhen parallel execution fails:\n- [ ] Check each subagent's output for errors\n- [ ] Verify no file conflicts between phases\n- [ ] Ensure dependencies were correctly identified\n- [ ] Check if task was too complex for parallelization\n- [ ] Review if phases had hidden dependencies\n- [ ] Consider falling back to sequential execution\n\n## Production Completion Criteria\n\nEach subagent must ensure its phase is **production-ready** before reporting complete:\n\n1. **All sub-tasks implemented**: Code written and committed\n2. **All tests pass**: Full test suite for the phase passes\n3. **End-to-end works**: The phase's functionality works in the full flow\n4. **No blocking issues**: Fix all issues discovered, don't defer them\n\n### Critical Rule for Subagents\n\nWhen a subagent encounters an issue during validation:\n\n- **DO NOT** report the phase as complete with caveats\n- **DO NOT** say \"phase done, but there's an unrelated issue\"\n- **DO** fix all blocking issues before reporting complete\n- **DO** re-validate end-to-end after each fix\n- **DO** only report complete when production-ready\n\n### Orchestrator Validation\n\nAfter all phases complete, the orchestrator must:\n\n1. Run full integration test suite across all phases\n2. Validate the complete end-to-end user flow\n3. Ensure all PRs are mergeable without conflicts\n4. If ANY issue found, assign to appropriate subagent to fix\n5. Only declare \"all phases complete\" when production-ready\n\n## AI Instructions\n\n1. **Analyze dependencies** before spawning subagents\n2. **Create execution batches** based on dependency graph\n3. **Spawn subagents** with `run_in_background: true`\n4. **Monitor progress** via `\u002Ftasks` command\n5. **Collect results** as subagents complete\n6. **Trigger dependent phases** when prerequisites met\n7. **Track PR status** for all parallel phases\n8. **Update task list** with real-time progress\n9. **Validate end-to-end** before declaring any phase complete\n10. **Fix all blocking issues** in-place, never defer them\n\n## References\n- See `reference.md`\n- Claude Code documentation on Task tool and subagents\n- See `.claude\u002Fskills\u002Ftask-processor-auto\u002FSKILL.md` for PR automation details\n\n---\n\n*Skill created for autonomous agent implementation with async subagents*\n","# Parallel Task Processor with Async Subagents\n\n## Prerequisites\n\n**Required:**\n- Claude Code v2.0.60+ (for async subagent support)\n- Git repository with remote configured\n- GitHub CLI (`gh`) installed and authenticated: `gh auth login`\n- GitHub repository (PR automation features require GitHub)\n\n**Optional:**\n- CI\u002FCD configured for automated checks\n- Sufficient API quota for parallel subagents\n\n**Non-GitHub Hosting:** If using GitLab, Bitbucket, or other platforms, the PR automation features won't work. You can still use parallel task processing, but create PRs manually.\n\n## Overview\nLeverages Claude Code's async subagents (v2.0.60+) to process multiple tasks in parallel, dramatically reducing implementation time.\n\n## Key Features\n- **Parallel Execution**: Spawn multiple subagents to work on independent tasks simultaneously\n- **Background Mode**: Subagents run in background, don't block orchestrator\n- **Auto-Notification**: Subagents notify orchestrator when complete\n- **PR Automation**: Automated gap analysis and PR comments after each phase\n\n## Parallelization Strategy\n\n### Level 1: Phase-Level Parallelization\nLaunch multiple phases in parallel when no dependencies exist. Use Claude Code's Task tool with multiple parallel calls:\n\n```\n\u002F\u002F Batch 1: Independent phases (run in parallel)\n\u002F\u002F Send a single message with multiple Task tool calls:\n\nTask(subagent_type: \"tdd-developer\", description: \"Phase 1: Comparison\", prompt: \"...\")\nTask(subagent_type: \"tdd-developer\", description: \"Phase 5: Performance\", prompt: \"...\")\nTask(subagent_type: \"tdd-developer\", description: \"Phase 8: Notifications\", prompt: \"...\")\n\n\u002F\u002F Wait for all to complete, then launch dependent phases\n```\n\n### Level 2: Sub-Task Parallelization\nWithin a single phase, parallelize independent sub-tasks:\n\n```\n\u002F\u002F Phase 1 split into 3 parallel streams (single message, multiple Task calls):\n\nTask(subagent_type: \"tdd-developer\", description: \"Discovery + Normalization\", prompt: \"Tasks 1.1, 1.2...\")\nTask(subagent_type: \"tdd-developer\", description: \"Ranking + Reasoning\", prompt: \"Tasks 1.3, 1.4...\")\nTask(subagent_type: \"tdd-developer\", description: \"Types + Tests\", prompt: \"Tasks 1.5, 1.6...\")\n\n\u002F\u002F When all complete → Final task\nTask(subagent_type: \"tdd-developer\", description: \"Create PR\", prompt: \"Task 1.7...\")\n```\n\n## Execution Protocol\n\n### 1. Analyze Dependencies\n```markdown\nPhase Dependencies:\n- Phase 1: None → Start immediately\n- Phase 2: Depends on Phase 1\n- Phase 3: Depends on Phase 2\n- Phase 4: Depends on Phase 3\n- Phase 5: None → Start immediately (parallel to 1-3)\n- Phase 6: Depends on Phase 5\n- Phase 7: Depends on Phase 3 + Phase 5\n- Phase 8: None → Start immediately (parallel to all)\n- Phase 9: Depends on all phases\n```\n\n### 2. Create Execution Batches\n```markdown\nBatch 1 (Parallel):\n- Phase 1 (large effort)\n- Phase 5 (medium effort)\n- Phase 8 (small effort)\n→ Total: Large effort (instead of 3x sequential!)\n\nBatch 2 (Parallel - after Phase 1):\n- Phase 2 (large effort)\n- Phase 6 (small effort, after Phase 5)\n→ Total: Large effort\n\nBatch 3 (Parallel - after Phase 2):\n- Phase 3 (large effort)\n→ Total: Large effort\n\nBatch 4 (Parallel - after Phase 3):\n- Phase 4 (medium effort)\n- Phase 7 (large effort, also needs Phase 5)\n→ Total: Large effort\n\nBatch 5 (Sequential - after all):\n- Phase 9 (medium effort)\n→ Total: Medium effort\n\n**Parallel execution reduces total effort by ~25% compared to sequential**\n```\n\n### 3. Spawn Async Subagents\n\n**Using Task Tool (parallel calls in single message):**\n\nTo launch multiple phases in parallel, send a single message with multiple Task tool calls:\n\n```\nTask tool call 1:\n  subagent_type: \"tdd-developer\"\n  description: \"Implement Phase 1: Comparison\"\n  prompt: \"Work through tasks 1.1 through 1.7 from tasks\u002F[task-file].md\n\n    Tasks:\n    - 1.1: Extend discovery engine\n    - 1.2: Create score normalization\n    - 1.3: Build ranking algorithm\n    - 1.4: Add comparison reasoning\n    - 1.5: Create TypeScript types\n    - 1.6: Integration tests\n    - 1.7: Create PR\n\n    Follow test-driven development.\n    Commit after each task.\n    When complete, create PR and run gap analysis.\"\n\nTask tool call 2 (in same message for parallel execution):\n  subagent_type: \"tdd-developer\"\n  description: \"Implement Phase 5: Performance\"\n  prompt: \"Work through tasks 5.1 through 5.7...\"\n\n\u002F\u002F Both subagents run in parallel\n\u002F\u002F Results returned when each completes\n```\n\n### 4. Monitor Progress\n\n**Check Subagent Status:**\n```bash\n# List running background tasks\n\u002Ftasks\n\n# Background tasks complete and return results automatically\n# Monitor the conversation for task completion notifications\n```\n\n### 5. Collect Results\n\nWhen a subagent completes, it notifies the orchestrator:\n```typescript\n\u002F\u002F Subagent 1 completes Phase 1\nResult: {\n  phase: 1,\n  pr_number: 16,\n  files_changed: 6,\n  tests_added: 50,\n  gap_analysis: \"docs\u002FPR_16_GAP_ANALYSIS.md\",\n  status: \"ready_for_review\"\n}\n\n\u002F\u002F Orchestrator receives notification\n\u002F\u002F Triggers next dependent phase (Phase 2)\n```\n\n### 6. Automated PR Workflow (Per Subagent)\n\nEach subagent follows this protocol when its phase completes:\n\n1. **Run Tests**: Full test suite for the phase\n2. **Create Branch**: `git checkout -b feat\u002Fphase-X-name`\n3. **Push**: `git push -u origin feat\u002Fphase-X-name`\n4. **Create PR**: Using `gh pr create`\n5. **Wait for CI**: 30-second delay\n6. **Check Status**: `gh pr view` + `gh pr checks`\n7. **Generate Gap Analysis**: Create `docs\u002FPR_X_GAP_ANALYSIS.md`\n8. **Push Gap Analysis**: Commit and push\n9. **Add PR Comment**: Detailed changes summary\n10. **Notify Orchestrator**: Return results\n\n## Example: Full Autonomous Run\n\n```bash\n# User starts the processor\nskill: task-processor-parallel\n\n# Orchestrator analyzes dependencies\n→ Identifies 5 execution batches\n→ Creates parallelization plan\n\n# Batch 1: Launch 3 phases in parallel\n→ Spawn Phase 1 subagent (tdd-developer)\n→ Spawn Phase 5 subagent (tdd-developer)\n→ Spawn Phase 8 subagent (tdd-developer)\n\n# Orchestrator: \"3 subagents launched, monitoring progress...\"\n\n# [After Phase 5 completes first - shortest effort]\n# Phase 5 completes first (shortest)\n→ Subagent 5 creates PR #20\n→ Subagent 5 runs gap analysis\n→ Subagent 5 notifies orchestrator: \"Phase 5 done, PR #20 ready\"\n\n# Phase 1 completes\n→ Subagent 1 creates PR #16\n→ Subagent 1 runs gap analysis\n→ Subagent 1 notifies orchestrator: \"Phase 1 done, PR #16 ready\"\n→ Orchestrator triggers Phase 2 (depends on Phase 1)\n\n# Phase 8 completes\n→ Subagent 8 creates PR #23\n→ Subagent 8 runs gap analysis\n→ Subagent 8 notifies orchestrator: \"Phase 8 done, PR #23 ready\"\n\n# Batch 2: Launch Phase 2 (depends on Phase 1 - just completed)\n→ Spawn Phase 2 subagent (tdd-developer)\n\n# ... Continue through all batches ...\n\n# All phases complete\n→ Orchestrator: \"All 9 PRs created and reviewed!\"\n→ User reviews PRs and merges\n```\n\n## Benefits\n\n### Time Savings\n- **Sequential**: Sum of all phase efforts\n- **Parallel**: Longest path through dependency graph\n- **Savings**: ~25% reduction in total effort\n\n### Efficiency\n- No idle time waiting for sequential tasks\n- Better resource utilization\n- Continuous progress across multiple fronts\n\n### Flexibility\n- Can pause\u002Fresume individual subagents\n- Can prioritize critical phases\n- Can adjust parallelization on the fly\n\n## Limitations & Considerations\n\n### Max Concurrent Subagents\nRecommended limits to avoid overwhelming resources:\n\n| Scenario | Max Parallel | Reason |\n|----------|--------------|--------|\n| Small tasks (\u003C 30 min each) | 4-5 | Token budget, context quality |\n| Medium tasks (1-2 hours) | 2-3 | Avoid merge conflicts |\n| Large tasks (half day+) | 1-2 | Resource contention |\n\n**Guidelines:**\n- Start conservative (2-3 parallel) and increase if stable\n- Monitor token usage with `\u002Fstats` command\n- If subagents produce poor results, reduce parallelism\n- Complex codebases benefit from less parallelism\n\n### Context Cost\n- Each subagent has its own context window\n- More parallel agents = higher token usage\n- Monitor costs with `\u002Fstats`\n\n### Merge Conflicts\n- Parallel work may touch same files\n- Mitigate: Design tasks to minimize overlap\n- Use clear file ownership per phase\n\n### CI Load\n- Multiple PRs may overload CI\n- Mitigate: Stagger PR creation\n- Use draft PRs to delay CI\n\n## Troubleshooting Parallel Subagents\n\n### Subagent Fails or Produces Poor Results\n\n**Symptoms:**\n- Subagent returns incomplete work\n- Code doesn't compile or tests fail\n- Subagent seems \"lost\" or unfocused\n\n**Diagnosis:**\n1. Check task complexity - may be too large for single subagent\n2. Check task dependencies - may need output from another phase\n3. Check context - subagent may lack necessary codebase knowledge\n\n**Solutions:**\n- Break task into smaller sub-tasks\n- Run the phase sequentially instead of parallel\n- Provide more context in the subagent prompt\n- Reduce number of parallel subagents\n\n### Subagent Hangs or Times Out\n\n**Symptoms:**\n- `\u002Ftasks` shows subagent running for unusually long time\n- No progress updates\n\n**Solutions:**\n1. Check `\u002Ftasks` for status\n2. If stuck, cancel and retry with smaller scope\n3. Consider running that phase sequentially\n\n### Merge Conflicts Between Parallel Phases\n\n**Symptoms:**\n- `git push` fails with conflict errors\n- PRs can't be merged\n\n**Solutions:**\n1. Identify which files conflict\n2. Determine which phase's changes are correct\n3. Manually resolve conflicts\n4. Re-run validation on merged code\n5. **Prevention:** Assign clear file ownership per phase\n\n### Inconsistent State Across Phases\n\n**Symptoms:**\n- Phase A created interface X\n- Phase B expected different interface\n- Integration fails\n\n**Solutions:**\n1. Define interfaces\u002Fcontracts BEFORE parallel execution\n2. Create shared types\u002Finterfaces in a prerequisite phase\n3. Re-run dependent phase after fixing interface\n\n### Debugging Checklist\n\nWhen parallel execution fails:\n- [ ] Check each subagent's output for errors\n- [ ] Verify no file conflicts between phases\n- [ ] Ensure dependencies were correctly identified\n- [ ] Check if task was too complex for parallelization\n- [ ] Review if phases had hidden dependencies\n- [ ] Consider falling back to sequential execution\n\n## Production Completion Criteria\n\nEach subagent must ensure its phase is **production-ready** before reporting complete:\n\n1. **All sub-tasks implemented**: Code written and committed\n2. **All tests pass**: Full test suite for the phase passes\n3. **End-to-end works**: The phase's functionality works in the full flow\n4. **No blocking issues**: Fix all issues discovered, don't defer them\n\n### Critical Rule for Subagents\n\nWhen a subagent encounters an issue during validation:\n\n- **DO NOT** report the phase as complete with caveats\n- **DO NOT** say \"phase done, but there's an unrelated issue\"\n- **DO** fix all blocking issues before reporting complete\n- **DO** re-validate end-to-end after each fix\n- **DO** only report complete when production-ready\n\n### Orchestrator Validation\n\nAfter all phases complete, the orchestrator must:\n\n1. Run full integration test suite across all phases\n2. Validate the complete end-to-end user flow\n3. Ensure all PRs are mergeable without conflicts\n4. If ANY issue found, assign to appropriate subagent to fix\n5. Only declare \"all phases complete\" when production-ready\n\n## AI Instructions\n\n1. **Analyze dependencies** before spawning subagents\n2. **Create execution batches** based on dependency graph\n3. **Spawn subagents** with `run_in_background: true`\n4. **Monitor progress** via `\u002Ftasks` command\n5. **Collect results** as subagents complete\n6. **Trigger dependent phases** when prerequisites met\n7. **Track PR status** for all parallel phases\n8. **Update task list** with real-time progress\n9. **Validate end-to-end** before declaring any phase complete\n10. **Fix all blocking issues** in-place, never defer them\n\n## References\n- See `reference.md`\n- Claude Code documentation on Task tool and subagents\n- See `.claude\u002Fskills\u002Ftask-processor-auto\u002FSKILL.md` for PR automation details\n\n---\n\n*Skill created for autonomous agent implementation with async subagents*\n",{"name":5,"category":23,"description":7},"main","47b89745588c3057f0ccd0517a2eb8967df27054","templates\u002F.claude\u002Fskills\u002Ftask-processor-parallel","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public","https:\u002F\u002Fgithub.com\u002Fdundas","derivative.io","2026-09-02 20:12:15",{"source":8,"skillName":5,"commands":42,"prompt":52,"zipAvailable":53},[43,46,49],{"tool":44,"command":45},"npx","npx skills add dundas\u002Fagentbootup-public --skill task-processor-parallel",{"tool":47,"command":48},"bunx","bunx skills add dundas\u002Fagentbootup-public --skill task-processor-parallel",{"tool":50,"command":51},"pnpm","pnpm dlx skills add dundas\u002Fagentbootup-public --skill task-processor-parallel","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Ftask-processor-parallel\u002FSKILL.md",true,[55,61,65,69],{"path":56,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":59,"sourceHtmlUrl":60},"catalog.json","REFERENCE","参考文档",980,"https:\u002F\u002Fgithub.com\u002Fjiayaoqijia\u002Fcryptoskill\u002Fblob\u002Fmain\u002Fskills\u002Ftrading\u002Fbankrbot-official-aeon-deal-flow\u002Fcatalog.json",{"path":62,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":63,"sourceHtmlUrl":64},"skill-bundle-manifest.json",846,"https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Ftask-processor-parallel\u002Fskill-bundle-manifest.json",{"path":66,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":67,"sourceHtmlUrl":68},"SOURCE.md",324,"https:\u002F\u002Fgithub.com\u002Fjiayaoqijia\u002Fcryptoskill\u002Fblob\u002Fmain\u002Fskills\u002Ftrading\u002Fbankrbot-official-aeon-deal-flow\u002FSOURCE.md",{"path":70,"resourceType":57,"resourceTypeDesc":58,"sizeBytes":71,"sourceHtmlUrl":72},"TRUST.auto.yaml",2217,"https:\u002F\u002Fgithub.com\u002Fjiayaoqijia\u002Fcryptoskill\u002Fblob\u002Fmain\u002Fskills\u002Ftrading\u002Fbankrbot-official-aeon-deal-flow\u002FTRUST.auto.yaml",[74,88,108,127,136,157],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":85,"publishedAt":29},"01M6CQZ7DFJB965BJJ5ZZ54CXY","fixture-skill","Secret-free recovery fixture skill.","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftests\u002Fruntime-adapters\u002Ffixtures\u002Fhermes\u002F0.18.2-darwin-arm64-real\u002Froot\u002Fskills\u002Ffixture-skill\u002FSKILL.md",[80],{"categoryNo":81,"categoryName":82,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000023","Skill与智能体开发","SCENARIO","使用场景",[86,87],"fixture skill","recovery",{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":99,"publishedAt":29},"01M6CQZ4NN0KVH1Z4E1Y1GJ3X9","user-story-generator","Generate standalone user stories for features without creating a full PRD. Perfect for backlog grooming and story refinement.","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Fuser-story-generator\u002FSKILL.md",[94],{"categoryNo":95,"categoryName":96,"taxonomyType":97,"taxonomyTypeDesc":98},"01JBSEEDCAT000000000000031","产品经理","PROFESSION","适用职业",[100,101,102,103,104,105,106,107],"user story generator","Story","user story","user","prd","documentation","generate","backlog",{"skillNo":109,"name":110,"nameZh":6,"description":111,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":112,"categories":113,"keywords":120,"publishedAt":29},"01M6CQZ16VC27HMWM31CCKK6BC","user-journey-mapper","Map user flows and journeys through your application for UX design and workflow visualization, independent of testing.","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Fuser-journey-mapper\u002FSKILL.md",[114,117],{"categoryNo":115,"categoryName":116,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000008","测试与质量",{"categoryNo":118,"categoryName":119,"taxonomyType":97,"taxonomyTypeDesc":98},"01JBSEEDCAT000000000000032","设计师",[121,122,103,105,123,124,23,125,126],"user journey mapper","journey","ux","visualization","design","testing",{"skillNo":128,"name":129,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":130,"categories":131,"keywords":132,"publishedAt":29},"01M6CQYXSQQJ4Z5CRAX2SSMHW1","transcript-query","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Ftranscript-query\u002FSKILL.md",[],[133,134,135],"transcript","transcript query","query",{"skillNo":137,"name":138,"nameZh":6,"description":139,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":140,"categories":141,"keywords":146,"publishedAt":29},"01M6CQYT7A1G4CPQTZKXR5TCK9","test-plan-generator","Generate comprehensive E2E test plans from PRDs, task lists, or descriptions with environment-aware prerequisites and issue tracking.","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Ftest-plan-generator\u002FSKILL.md",[142,143],{"categoryNo":115,"categoryName":116,"taxonomyType":83,"taxonomyTypeDesc":84},{"categoryNo":144,"categoryName":145,"taxonomyType":83,"taxonomyTypeDesc":84},"01JBSEEDCAT000000000000010","后端与接口",[147,148,149,150,21,151,152,153,154,155,106,156],"plan","test plan generator","test plan","test","issue","e2e test","environment","tracking","quality","e2e",{"skillNo":158,"name":159,"nameZh":6,"description":160,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":161,"categories":162,"keywords":164,"publishedAt":29},"01M6CQYQAWFSPTYE2VTJJVW72G","tasklist-generator","Generate high-level tasks and gated sub-tasks from a PRD, with agent assignments, commit\u002FPR strategy, and relevant files.","https:\u002F\u002Fgithub.com\u002Fdundas\u002Fagentbootup-public\u002Fblob\u002Fmain\u002Ftemplates\u002F.claude\u002Fskills\u002Ftasklist-generator\u002FSKILL.md",[163],{"categoryNo":81,"categoryName":82,"taxonomyType":83,"taxonomyTypeDesc":84},[165,166,167,168,104,106,24,169,170,23],"tasklist generator","gated","files","strategy","pr","commit"]