sakura-plan

Converts accepted software outcomes into executable development plans with dependencies and checkpoints.

3|1|Updated Sep 9, 2026
One-click install
npx skills add https://github.com/v20227/using-sakura-dev --skill sakura-plan-v20227
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sakura-plan
Source: https://github.com/v20227/using-sakura-dev/tree/main/skills/sakura-plan
Command: npx skills add https://github.com/v20227/using-sakura-dev --skill sakura-plan-v20227

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Turning an agreed product goal into an ordered, recoverable set of development tasks is hard when work spans multiple modules, dependencies, and integration points. This Skill structures that transition so implementation starts from a verified, sequenced plan instead of ad-hoc edits. ## Core Features & Use Cases - Executable Work Decomposition: Gives every planned item an identity, concrete result, owning area, prerequisite IDs, acceptance reference, and current state. - Dependency-Aware Sequencing: Orders items by prerequisites and user value, selects one ready item, and preserves completed work and failed approaches when revising the plan. - Recovery and Continuity: Maintains branches, worktrees, checkpoints, and recovery details so the plan survives context switches and resumed sessions. - Use Case: You have an accepted feature spanning frontend, API contract, and database changes. Use this Skill to split it into vertical slices with clear ownership and verification, then begin implementation on the first ready item. ## Quick Start Ask Codex to break the accepted feature into an ordered development plan with dependencies, owners, and verification steps before writing any code.

Frequently Asked Questions about sakura-plan

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

FAQPage Schema
How do I turn a feature request into an executable development plan?▼

Start from an accepted outcome and current project facts, then decompose the work into items each having an identity, concrete result, owning area, prerequisites, and acceptance reference. Order items by dependencies and user value, then select one ready item to implement first.

When should I create a development plan instead of editing code directly?▼

Plan when work spans multiple meaningful changes, dependencies, integration points, or recovery checkpoints. Do not use planning merely to restate a simple edit; independent small modifications should stay short and direct.

How do I handle cross-boundary features spanning frontend, backend, and database?▼

Identify the applicable frontend, contract, backend, database, integration, and end-to-end work for the feature. Define the owning module, entrypoint, dependency direction, public contract, and data owner before implementation, and only explain omissions when they could change the accepted result.

Can a development plan be revised after implementation starts?▼

Yes. When revising, preserve completed items and failed approaches, and update the same accepted workstream for added requests unless they change the milestone. Correct any plan assumptions that conflict with repository or runtime evidence.

What are the limitations of plan-driven development workflows?▼

A plan should only cover enough to make the next result executable, verifiable, and recoverable; expanding it beyond that adds no value. It also must not include unauthorized commits, pushes, deployments, credential use, or other external mutations.