🌿 lifeplan v1.0.0 · MIT

A pragmatic SQLite-backed task manager built for AI/human shared work queues — priority, a verification-first lifecycle, waiting-for tracking, triage, and a soft archive that never hard-deletes.

What it does

Verification-first lifecycle: active → done_untested → done_tested → completed — marking work complete and knowing it works are different events. Verified tasks graduate automatically after a 14-day cooling-off.
Waiting-for tracking: record who a task is blocked on and why; triage surfaces stale waits ("is this still needed?").
Triage report: stale actives (>7d), review-stale waits (>14d), done-but-untested items (>3d), archive candidates (>30d).
Soft archive: archive --days N moves finished rows to a tasks_archive table — full content preserved, nothing hard-deleted.
Full audit trail: every create/complete/priority/status change logged to task_history.
LLM-aware fields (optional): --model, --complexity for AI-driven queues.
Single file, zero dependencies — Python stdlib only.

Install (pip)

pip install git+https://mandrilly.com/git/lifeplan.git

Installs a lifeplan command. Database location (first match wins): --db PATH · $LIFEPLAN_DB · ~/.lifeplan/tasks.db.

Quick start

lifeplan add --title "Draft landing page" --beneficiary self --priority 10 --domain work
lifeplan list --status active --limit 5
lifeplan complete --id task-abcd1234 --result "Published"      # → done_untested
lifeplan update --id task-abcd1234 --status done_tested        # after verifying
lifeplan update --id task-abcd1234 --waiting-for "client approval"
lifeplan triage                                                # daily hygiene report
lifeplan archive --days 30                                     # soft-move to tasks_archive

Source

git clone https://mandrilly.com/git/lifeplan.git

Requires Python 3.9+. MIT license.