Conciseness

The body is dense and imperative, assumes competence (no explanations of zod, drizzle, transactions, or hooks), and every rule is non-obvious project-specific knowledge (e.g., 'Plain `pnpm doctor` runs pnpm's own command and exits 0', the 6,000-character AGENTS.md cutoff). Nothing is padded, matching anchor 5 rather than 4.

Actionability

Fully executable guidance throughout: copy-paste commands ('pnpm exec agent-native dev', 'pnpm action <name> --key value'), concrete code patterns ('export * from "@agent-native/toolkit/ui/dialog";', 'fail(message, { errorCode, statusCode })' with 409/404/403 mapping, 'toast.error(actionErrorMessage(error) ?? "<fallback>")'), and exact file paths. Matches anchor 5's copy-paste-ready coverage of common cases.

Workflow Clarity

Five clearly sequenced phases (Orient, Decide, Build, Verify, Report) with six numbered up-front decisions and an explicit ordered verification sequence including feedback loops ('Fix every finding, then rerun only what failed', guarded write called twice, browser smoke asserting the rejection toast). Destructive database operations are guarded (additive migrations, 'Never edit an applied entry'), so the workflow-clarity cap does not apply; this matches anchor 5.

Progressive Disclosure

The body holds the decisions and core workflow while bulk detail is split across five real, one-level-deep reference files (edit-points-chat/default selected by scaffold template, data-and-access, rooms-example, verification), each clearly signaled by path and verified to exist with no nested references. This matches anchor 5's well-signaled, appropriately split structure.

Specificity

The what-clause ('turning this starter into a working domain app from a short or vague request') names the domain and one concrete action, but does not enumerate several distinct capabilities. It fits anchor 3 rather than 4 because the verb list (build, make, create, prototype) serves as trigger phrasing, not a list of what the skill does.

Completeness

It explicitly answers what ('Fast end-to-end workflow for turning this starter into a working domain app') and when ('Use when asked to build, make, create, or prototype...'), with concrete trigger phrases plus 'Read before the first edit' as additional trigger guidance. This matches anchor 5 exactly; anchor 4's weaker 'when' does not apply.

Trigger Term Quality

'build, make, create, or prototype an app, tool, tracker, dashboard, booking, inventory, or CRM system' comprehensively covers the natural verbs and object nouns users say when requesting this work, with synonyms. No meaningful natural terms are missing for this domain, so it matches anchor 5 rather than 4.

Distinctiveness Conflict Risk

'this starter' and 'with Agent-Native' scope it to a clear niche, but generic trigger nouns like 'tool' and 'dashboard' create minor overlap risk with closely related skills (e.g., charting or feature-addition skills). This places it at anchor 4 rather than 5's 'minimal conflict risk'.