Build Task Parser

Parses tasks.md to extract structured task information for build orchestration.

CRITICAL: Output ONLY JSON. No explanations, no progress updates.

0. Parameters

Name	Position	Default	Purpose
TASKS_PATH	$1	(required)	Path to tasks.md file

<tasks_path>$1</tasks_path>

1. Parsing Algorithm

1.1 Read File

Read the file at {TASKS_PATH}.

1.2 Extract Tasks

Use this regex pattern on each line:

- \[([ x!])\] \*\*([^*]+)\*\*: (.+?)(?:\s*`\[complexity:(simple|medium|complex)\]`)?$

Capture groups:

Group	Field	Values
1	status	space=pending, x=done, !=blocked
2	task_id	T1, T1.1, TD1, etc.
3	description	Task text
4	complexity	simple/medium/complex (default: medium)

1.3 Categorize Tasks

• implementation_tasks: task_id starts with T (not TD)
• doc_tasks: task_id starts with TD

1.4 Error Handling

• Malformed entries: Log in warnings array, skip
• Missing complexity: Default to medium
• Empty file: Return empty arrays

2. Output Contract

Return ONLY this JSON:

{
  "status": "success",
  "implementation_tasks": [
    {"task_id": "T1", "status": "pending", "description": "Create feature", "complexity": "medium"},
    {"task_id": "T2", "status": "done", "description": "Add tests", "complexity": "simple"}
  ],
  "doc_tasks": [
    {"task_id": "TD1", "status": "pending", "description": "Update docs", "complexity": "simple"}
  ],
  "summary": {
    "total": 3,
    "pending": 2,
    "done": 1,
    "blocked": 0
  },
  "warnings": []
}

Fields:

• implementation_tasks: Array of T* tasks
• doc_tasks: Array of TD* tasks
• summary: Counts by status
• warnings: Any parsing issues encountered

3. Anti-Loop

EXECUTE IMMEDIATELY:

• Do NOT ask for clarification
• Execute once, output JSON, STOP
• No iteration or refinement

4. Output Discipline

CRITICAL - Silent Execution:

• Do ALL work in <thinking> tags
• Output ONLY the final JSON
• No progress updates, no explanations