reversa-to-do

Decomposes a feature roadmap into atomic tasks with IDs, dependencies, and parallelism markers.

Updated Jun 26, 2026
One-click install
npx skills add https://github.com/RLuf/fzagent --skill reversa-to-do-rluf
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reversa-to-do
Source: https://github.com/RLuf/fzagent/tree/main/.agents/skills/reversa-to-do
Command: npx skills add https://github.com/RLuf/fzagent --skill reversa-to-do-rluf

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Turning a high-level roadmap into an executable task list is error-prone: tasks end up too large, dependencies are unclear, and parallel work opportunities are missed. This Skill converts a roadmap.md file into a structured actions.md with atomic, trackable tasks. ## Core Features & Use Cases - Atomic decomposition: Breaks each roadmap item into tasks executable in a single agent turn, split across five standard phases (preparation, tests, core, integration, polish). - Stable IDs and dependency tracking: Assigns zero-padded IDs (T001, T002...) that are never recycled, with explicit dependency columns and a [//] marker for parallelizable tasks. - Confidence inheritance: Carries the green/yellow/red confidence markers from roadmap decisions into each task row. - Use Case: After running /reversa-plan to produce a roadmap for a payments feature, invoke this Skill to generate an actions.md table with 25 tasks, 8 marked parallelizable, ready for /reversa-coding. ## Quick Start Type /reversa-to-do to decompose the current feature roadmap into an executable actions.md task list.

Frequently Asked Questions about reversa-to-do

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I break a software roadmap into atomic tasks?▼

Run the reversa-to-do skill after generating a roadmap.md with /reversa-plan. It decomposes each roadmap item into tasks small enough to complete in one agent turn, organized across five phases: preparation, tests, core, integration, and polish.

How to mark tasks that can run in parallel in a task list?▼

Tasks that touch different files and have no mutual dependencies are marked with [//] at the start of the line. The generated actions.md also includes a dependencies column listing blocking task IDs and a summary of total parallelizable actions.

What happens if roadmap.md is missing when running reversa-to-do?▼

The skill aborts with a clear message pointing to /reversa-plan, since the roadmap is a required input. It also checks for active-requirements.json first and directs you to /reversa-requirements if that is absent.

When should a task be split into smaller tasks?▼

Split a task when it has more than five logical subpoints, touches more than three unrelated files, or contains sequencing words like 'and also' or 'then'. A task is atomic only if an agent can finish it in one turn without human feedback.

Are task IDs reused after a task is removed?▼

No. IDs are never recycled, even if a task is removed in a later review. Renumbering only happens when the actions.md document is generated for the first time.