groom-backloglisted

# Groom Backlog Skill Grooms the local task database by identifying completed, redundant, incorrectly prioritized, or unassigned tasks. > Use `/create-task` for task creation — handles decomposition, deduplication, criteria, and deps. Use `tusk task-insert` only for bulk/automated inserts. ## Step 0: Start Cost Tracking Record the start of this groom run so cost can be captured at the end: ```bash tusk skill-run start groom-backlog ``` This prints `{"run_id": N, "started_at": "..."}`. Capture `run_id` — you will need it in Step 7. > **Early-exit cleanup:** If any check below causes the skill to stop before Step 7b (e.g., `tusk setup` / `tusk backlog-scan` fails, the backlog is empty with nothing to groom, or the user declines the Step 4 approval prompt), first call `tusk skill-run cancel <run_id>` to close the open row, then stop. Otherwise the row lingers as `(open)` in `tusk skill-run list` forever. ## Setup: Fetch Config and Backlog Before grooming, fetch everything needed in a single call: ```bash tusk setup ``` This returns a JSON object with two keys: - **`config`** — full project config (domains, agents, task_types, priorities, complexity, etc.). Use these values (not hardcoded ones) throughout the grooming process. - **`backlog`** — all open tasks as an array of objects. Use this as the primary backlog data for Step 1 (you still need the dependency queries below). ## Pre-Check: Auto-Close Stale Tasks Run the auto-close check for moot contingent tasks: ``