lightweight-task-workflow

To maintain task context for an AI agent across multiple sessions, this Skill uses persistent state files like `tasks.md`, `requirements.md`, and `session.md` to track progress and ensure continuity. It guides the AI through a strict state machine for predictable execution.

A state machine workflow for multi-session task management is a defined sequence of states like CHECK_STATUS, WORKING, and VERIFY that guides an AI agent through task execution. This enforces predictable steps and requires verification before completion.

To prevent loss of context on complex software features spanning multiple days, this Skill persists task lists and session state in dedicated files. It ensures the AI remembers exactly where it left off and what needs to be done next.

Yes, you can use this task management workflow for complex projects requiring continuity and detailed progress tracking. It maintains persistent state in `.claude/` files to track requirements and session progress across different work periods.

No, you do not need any external dependencies to run this persistent task workflow. The Skill operates independently by storing task lists, requirements, and session state directly in local files.

The best way to verify task completion using an AI state machine is to enforce a dedicated VERIFY state. This Skill mandates verification steps within its workflow sequence before allowing a task to be marked as complete.