ac-task-executor（ac 任务执行器）

AC Task Executor

Execute implementation tasks for autonomous coding features.

Purpose

Orchestrates the execution of feature implementations, managing the task queue and coordinating between planning, implementation, and verification phases.

Quick Start

from scripts.task_executor import TaskExecutor

executor = TaskExecutor(project_dir)
result = await executor.execute_next_feature()

Execution Modes

Single Feature

Execute one feature at a time:

result = await executor.execute_feature("auth-001")

Queue Mode

Process feature queue continuously:

await executor.execute_queue(max_features=10)

Batch Mode

Execute multiple features in a batch:

results = await executor.execute_batch(["auth-001", "auth-002"])

Task Lifecycle

1. SELECT   → Choose next feature from queue
2. PREPARE  → Load context, generate tests (RED)
3. IMPLEMENT → Write code to pass tests (GREEN)
4. REFACTOR → Clean up while tests pass
5. VALIDATE → Run criteria validation
6. COMMIT   → Git commit changes
7. UPDATE   → Mark feature as passes: true
8. NEXT     → Continue to next feature

Execution Result

{
  "feature_id": "auth-001",
  "status": "completed",
  "phases": {
    "prepare": {"success": true, "duration_ms": 1500},
    "implement": {"success": true, "duration_ms": 45000},
    "validate": {"success": true, "duration_ms": 5000},
    "commit": {"success": true, "commit_hash": "abc123"}
  },
  "tests": {
    "total": 5,
    "passed": 5,
    "failed": 0
  },
  "metrics": {
    "tokens_used": 15000,
    "estimated_cost": 0.23
  }
}

Error Handling

• Validation Failure: Retry implementation up to 3 times
• Test Failure: Analyze failures, adjust approach
• Timeout: Save state, create handoff for continuation
• Critical Error: Rollback to last checkpoint

Integration

• Uses: ac-state-tracker, ac-test-generator, ac-criteria-validator
• Triggers: ac-commit-manager, ac-qa-reviewer
• Reports to: ac-session-manager

API Reference

See scripts/task_executor.py for full implementation.