Iterate

Iterate: Multi-Agent Iteration Orchestration

Execute a complete iteration cycle following the file-based orchestration approach with Planning Petra coordinating specialist agents through TDD/ATDD workflow.

Instructions

Acting as orchestrator in file-based agent collaboration, execute a batch:

1. Planning Petra: Assess current state

   • Read work/collaboration/inbox/ for pending tasks
   • Identify next batch (check NEXT_BATCH.md if exists)
   • Priority order: critical > high > medium > low

2. Execute batch (specialist agents):

   For each task in batch (priority descending):

   a. Initialize as assigned agent (backend-dev, architect, frontend, etc.)

   b. Start task using script:

   python tools/scripts/start_task.py TASK_ID
   

   c. Follow TDD/ATDD cycle:

   • RED: Write failing test first
   • GREEN: Implement minimum code to pass
   • REFACTOR: Improve code quality

   d. Run all tests: ensure passing before proceeding

   e. Complete task using script:

   python tools/scripts/complete_task.py TASK_ID
   

   (Validates result block exists and moves to done/{agent}/)

   f. Create work log (directive 014) in work/reports/logs/<agent>/

   If blocked: Use freeze script instead:

   python tools/scripts/freeze_task.py TASK_ID --reason "Specific blocking reason"
   

3. Planning Petra: Update artifacts

   • Update roadmap if exists (docs/architecture/roadmap-*.md)
   • Update NEXT_BATCH.md with next priorities
   • Document blockers or decisions

4. Provide executive summary:

   • Tasks completed (time: estimated vs actual)
   • Tests passing (coverage %)
   • Decisions made or deferred
   • Blockers identified
   • Next recommended batch

5. Review gate (conditional):

   • If significant architectural changes: create Alphonso review task
   • If routine implementation: ready for next iteration

Quality Gates

Before marking batch complete:

• ✅ All tests passing
• ✅ Coverage >80% (target)
• ✅ Work logs created (directive 014)
• ✅ Tasks completed using complete_task.py script (moved to done/<agent>/)
• ✅ Planning artifacts updated
• ✅ Any blocked tasks frozen using freeze_task.py

Directives Applied

• Directive 014: Work Log Creation
• Directive 016: Acceptance Test-Driven Development (ATDD)
• Directive 017: Test-Driven Development (TDD)
• Directive 018: Traceable Decisions (ADRs if needed)
• Directive 019: File-Based Orchestration

Agent Handoff Pattern

Planning Petra
    ↓ (identifies batch)
Specialist Agent(s)
    ↓ (executes with TDD)
Planning Petra
    ↓ (updates artifacts)
Architect Alphonso (if needed)
    ↓ (reviews & approves)
[Ready for next /iterate]

Example Usage

User: /iterate

Agent (Petra): Starting M2 Batch 2.3 (2 tasks: GenericYAMLAdapter + ENV support)...

Agent (Backend-Dev):
  - Task 1: GenericYAMLAdapter [HIGH]
    - Started: `python tools/scripts/start_task.py 2026-02-05-generic-yaml-adapter`
    - RED: 24 failing tests written
    - GREEN: Implementation complete, all tests pass
    - REFACTOR: Extracted common patterns
    - Completed: `python tools/scripts/complete_task.py 2026-02-05-generic-yaml-adapter`
    - Result: 82% coverage, 3.5h (vs 5h estimated)

  - Task 2: ENV variable support [HIGH]
    - Started: `python tools/scripts/start_task.py 2026-02-05-env-support`
    - RED: 20 failing tests written
    - GREEN: Template expansion implemented
    - REFACTOR: Security validations added
    - Completed: `python tools/scripts/complete_task.py 2026-02-05-env-support`
    - Result: 100% coverage, 1.5h (vs 2h estimated)

Agent (Petra):
  ✅ M2 Batch 2.3 COMPLETE (2/2 tasks, 5h actual vs 7h estimated)
  ✅ 44 tests passing (92% coverage)
  ✅ Work logs: work/reports/logs/backend-dev/2026-02-05-m2-batch-2.3.md
  ✅ Roadmap updated: M2 Tool Integration 100% complete
  ✅ Tasks properly completed using scripts (validation passed)
  📊 Overall progress: 45% (M1-M2 done, M3-M4 remaining)

  Next recommended batch: M3 Batch 3.1 - Telemetry Infrastructure (3 tasks)

  Review gate: Significant architectural change (generic adapter approach)
  ✅ Created task: work/collaboration/inbox/2026-02-05-architect-m2-review.yaml

Related Skills

• /status - Check current planning state before iterating
• Use agents/prompts/iteration-orchestration.md for detailed reference and advanced workflows

References

• Prompt Template: agents/prompts/iteration-orchestration.md (comprehensive reference)
• Approach: .github/agents/approaches/work-directory-orchestration.md
• Directive 014: .github/agents/directives/014_work_log_creation.md
• Directive 016: .github/agents/directives/016_acceptance_test_driven_development.md
• Directive 017: .github/agents/directives/017_test_driven_development.md
• Directive 019: .github/agents/directives/019_file_based_collaboration.md