- name
- groove-utilities-task-analyse
- description
- Analyse task status and distribution across the configured backend.
- license
- MIT
- allowed-tools
- Read Write Edit Glob Grep Bash(git:*) Bash(beans:*) Bash(gh:*) Bash(linear:*) Bash(npx:*) AskUserQuestion
- metadata
- {"author":"andreadellacorte"}
# groove-utilities-task-analyse
## Outcome
Tasks are summarised by status. Milestones and epics are summarised. Completed tasks include enough resolution detail for memory population bullets. Output is suitable for daily end and daily memory.
## Acceptance Criteria
- Tasks grouped by status: in-progress, todo/ready, blocked, completed, scrapped
- Milestones/epics show progress (X of Y tasks done)
- Completed tasks include resolution summary (not just title)
- Output contains enough detail to write meaningful daily memory bullets without re-reading tasks
## Constraints
- Read `tasks.backend` and `tasks.analyse_limit` (default 30) from `.groove/index.md`
- If `tasks.backend: none`, no-op with friendly message
- Backend mappings:
- `beans`: (1) run `skills/groove-utilities-task-list/scripts/list-tasks-by-priority.sh [LIMIT]` for active tasks (LIMIT from `tasks.analyse_limit` or 30); (2) run `beans list --json --status completed` and `beans list --json --status scrapped` for completed/scrapped; merge, group by status, then summarise (milestones/epics and resolution detail as per acceptance criteria)
- `linear`: fetch all assigned issues via linear CLI or MCP, group by state
- `github`: `gh issue list --state all --assignee @me`, group by label/state
- For completed tasks: include the body's "Summary of Changes" section if present
- If a completed task has no resolution, note that it is missing — do not omit the task
- Output format should be scannable markdown (headers per status group, bullet per task)
在 GitHub 查看