reverse-to-do

To convert a roadmap into executable actions, decompose roadmap.md into an actions.md file containing atomic tasks, stable sequential IDs, dependencies, and parallelism indicators for agent execution. This ensures reliable autonomous execution of your feature plan.

The best way to break down a product roadmap into atomic tasks is to decompose each roadmap item into one-turn, coherent actions suitable for autonomous execution. This process assigns deterministic IDs and records dependencies and confidence levels for stable execution planning.

You mark tasks for parallel execution in an action plan by using [//] indicators when tasks touch different files and do not depend on each other. This parallelism signaling allows concurrent processing during autonomous agent execution.

Yes, you can generate an actions.md file directly from a roadmap.md document by applying standard before-to-do and after-to-do hooks. This governed output generation produces feature-dir/actions.md with atomic tasks, IDs, dependencies, and parallelism markers.

Yes, you need specific state validation before decomposing a roadmap into tasks. The process requires reading Reversa state for output paths, validating active requirements, loading optional delta or interface context, and applying before-to-do and after-to-do hooks.

Stable sequential IDs like T001 and T002 are used for task decomposition in agent workflows to ensure deterministic execution planning. They record dependencies and confidence levels, which are critical for maintaining order and reliability during autonomous execution.