Skip to main content
Reference — generated from the toolkit

Source of truth: toolkit/scaffold/CLAUDE.md. Edit it there; this page is regenerated on build.

{Product name}

A RAPID-built product. AI agents build; engineers direct, review and govern.

What this is​

One paragraph: what the product does, who it's for, its place in the portfolio.

Key commands​

{install}      # …
{dev}          # run locally
{test}         # run the test suite (unit + mutation)
{lint}         # format + lint

Architecture​

Pointers, not prose: where the entry points are, the main modules, the data model (docs/), the reference architecture it maps to. Link to ADRs in docs/adr/ for why — do not restate decisions here.

Conventions & standards that apply​

  • Build to the Definition of Done and coding standards; a human always merges.
  • Read memory/ (gotchas) and the relevant runbook at session start.
  • Never take an irreversible action — prepare it and stop for a named human.
  • No new dependency without explicit approval in the PR.

Session start checklist for agents​

  1. Read this file, memory/, and the runbook for the area you're touching.
  2. Confirm the task brief, acceptance criteria and files in scope.
  3. Build → run tests/lint → open a PR with the AI-provenance declaration.