Edition Detection for AI Workflows

This page applies the general detection rule from Editions / Detection to AI-assisted development specifically.

Edition detection should happen before any AI workflow that assumes a frontend stack, module set, or example path.

AI-specific consequence

If an agent skips edition detection, it can:

scaffold against the wrong UI-layer assumptions
suggest the wrong frontend flavors
cite suites, modules, SSR site baselines, or project assets that do not exist in the active repo
generate docs or skills that accidentally hardcode Basic-only behavior or npm create cabloy assumptions into Start workflows

Operational rule

Before an AI workflow recommends implementation steps:

check the edition marker
verify the relevant package scripts or CLI entrypoints
branch the guidance only where the editions truly diverge

Where this rule should live

Keep the same rule in three places on purpose:

public docs so people can review it
skills so agents execute it consistently
CLAUDE.md so repo-wide Claude behavior stays aligned