backlog-start-task

Purpose

Start work on an existing backlog task by:

1. updating the standardized header table (status, started_on), and
2. creating a todowrite checklist for execution based on the task’s subtasks.

Prerequisites

• The task file exists in .backlog/ and matches TXXX-*.md.

Procedure

Use todowrite with the below checklist.

1. Identify the task

   • Ask the human for the task ID (TXXX) or the exact filename.
   • Locate the file in .backlog/ (NOT in .backlog/.archive/).

2. Update task header table

   • Update the standardized header table (immediately below the title):
     • set status to IN PROGRESS
     • set started_on to YYYY-MM-DD HH:mm
   • If the header table is missing, add it.
   • If started_on is already set, ask the human whether to keep it or overwrite it.

3. Create the working checklist (todowrite)

   • Parse the task file for an optional ## Subtasks section.
   • If subtasks exist:
     • create one todo per subtask (use the subtask ID like TXXX-01 in the todo content).
   • If no subtasks exist:
     • create a single todo for the main task (TXXX).
   • Set exactly one item to in_progress (typically the first actionable item).
   • All other items should be pending.

4. Verify

   • The task file’s header table has:
     • status = IN PROGRESS
     • started_on = YYYY-MM-DD HH:mm
   • The todowrite list mirrors the subtasks (or the main task if none).

Notes

• This skill does not move files; archiving is handled by backlog-archive-task.