SuperAPI AI Directory

rewrite-plan

Testing & quality

View repository

Updated 10-16 09:34

About this skill

Rewrite plan

Spec §20.3 / §21.3.2: scenario 4 (design → deliverable production code) requires a contract-shaped rewrite plan before the agent starts editing files. The plan is the audit trail every subsequent stage references; without it the patch path slides into "refactor everything that looks wrong" and the build breaks in unreviewable ways.

Inputs

  • code/index.json from code-import.
  • token-map/*.json from token-map.
  • The active design system DESIGN.md (already in prompt).
  • The user's tune intent (typically a short brief).

Output

project-cwd/
└── plan/
    ├── plan.md           # human-readable narrative
    ├── ownership.json    # { file: '...', layer: 'leaf' | 'shared' | 'route' | 'shell' }
    ├── steps.json        # ordered { id, files[], rationale, risk: 'low' | 'medium' | 'high' }[]
    └── meta.json         # { generatedAt, atomDigest, tokenMapDigest }

steps.json is the input patch-edit reads one entry per iteration. ownership.json is the source of truth for "which files belong to a single component vs. shared infrastructure"; patch-edit refuses to touch a shell-tier file unless the matching step has risk: 'high' and the user explicitly confirmed.

Convergence

The atom completes when plan.md is non-empty AND steps.json contains at least one step.

Anti-patterns the prompt fragment forbids

  • Plans that include "rewrite the whole component library" as a single step. Break into per-component steps.
  • Plans that don't classify ownership.
  • Plans that don't list a build-test step at the end.

Status

Reserved id, prompt-only fragment in v1. The component-mapping heuristic + ownership classifier land in spec §16 Phase 7.