task-system
Installation
SKILL.md
AgentCLI Task System
A lightweight, static, file-based task management system embedded directly in the repository. Tasks are permanent history — once created, they are never deleted. This system exists to help AI agents and humans manage project complexity with a simple CLI that needs no database, no server, and no internet.
This is separate from any runtime/ephemeral task tools you may have. Those are temporary — they disappear when the session ends. This task system lives in the repository. Its tasks are kept forever.
Core Philosophy
- Static, file-based, in-repo — Three files in
tasks/:inbox.md,tasks.yaml,closed.yaml. No database, no service, no setup beyondtasks init. - Tasks are permanent history — Nothing is ever deleted. Tasks flow from
tasks.yaml(active) toclosed.yaml(archive). Once a task is created, its record persists for the life of the repository. - Purpose: manage project complexity — Track what needs doing, what is in progress, what is done. Keep the agent and human aligned on priorities without losing context across sessions.
- 12 statuses, any lifecycle — Tasks can hold any of 12 statuses:
todo,in-progress,done,blocked,postponed,cancelled,review,waiting,parked,deferred,backlog,abandoned. You move between them freely withtasks update --status <name>— there are no restrictions on transitions. - Counter-based IDs — Tasks get
TSK-NNNNIDs (e.g.TSK-0001). Auto-incrementing, no gaps are guaranteed (skipped IDs on concurrent writes are acceptable). - Newest-first ordering — Recent tasks appear at the top of the active list. Closed tasks are oldest-first (chronological append).
- Growing complexity only when needed — The system intentionally started minimal. New features are added only when a real need emerges.