Conciseness

The body is efficient: an 8-step summary delegates the full workflow to a canonical doc, the error table is compact, and the React 19 / @portaljs/ckan caveat is genuinely non-obvious knowledge. Minor over-explanation could be trimmed — e.g., "so the catalog is pre-rendered at build time and the site can still be statically deployed" restates the point already made by "fetch CKAN server-side in getStaticProps/getStaticPaths", and the Overview repeats the catalog/showcase rewire story told again in Instructions and Output. Fits anchor 4 (efficient, minor instances that could be trimmed) rather than 5.

Actionability

Concrete and mostly executable: named files (lib/ckan.ts, pages/search.tsx, pages/[owner]/[slug].tsx), named endpoints (package_search?rows=1, organization_show), runnable commands (npx next build, npm run i18n:check), and copy-paste invocations in Examples. The gap is that the main artifact — the generated fetch client — is specified but not shown; the actual code lives in the external canonical doc, so the body alone gives a specification rather than code. That is a minor gap, matching anchor 4 rather than 5; it is well above anchor 3 since steps, commands, and validation calls are all concrete.

Workflow Clarity

Eight clearly sequenced steps with explicit validation checkpoints: step 3 verifies the CKAN backend is reachable and validates org filters, step 7 verifies the build ("fix any error before reporting success"), and step 8 reports what changed. The Error Handling table adds feedback loops (re-prompt on bad URL, list valid orgs via organization_list, fix the first error and re-run the build). This matches the anchor-5 pattern of sequence + explicit validation + error-recovery loops; anchor 4 would require a missing checkpoint.

Progressive Disclosure

Well-organized sections (Overview, Prerequisites, Instructions, Output, Error Handling, Examples, Resources) with one-level-deep references that all resolve: references/reference.md exists in the bundle and holds the detail (env vars, endpoint table, filters, troubleshooting) without nesting further, and the canonical workflow doc is clearly signaled in both Instructions and Resources. Matches anchor 5 (clear overview, well-signaled one-level references, easy navigation); not 4 since there are no organization gaps to cite.

Specificity

Lists multiple concrete actions — "Wire a scaffolded PortalJS portal to a CKAN backend over its API", "Generates a tiny server-side fetch client (no runtime dependency)", "feeds the /search catalog and /@namespace/slug showcases from CKAN instead of datasets.json" — naming the exact routes, artifacts, and what gets replaced. All verbs are third person and every action is concrete; this matches the comprehensive-coverage anchor rather than the score-4 anchor, which permits minor gaps.

Completeness

Explicitly answers both questions: what ("Generates a tiny server-side fetch client… feeds the /search catalog and /@namespace/slug showcases from CKAN") and when ("Use when connecting an existing portal to a live CKAN instance instead of a static manifest"). The "Use when" clause is present, specific, and includes concrete trigger conditions, matching the anchor-5 example pattern; a 4 would require the when-clause to be less explicit than it is.

Trigger Term Quality

Includes natural terms a user would say — "connecting", "CKAN", "portal", "live CKAN instance", "backend", "static manifest", "datasets.json" — but coverage is not exhaustive: common variations like "data portal", "link/hook up", or "catalog backend" phrasings are absent. Good keyword coverage with a few natural terms missing fits anchor 4; not 3 (the key phrase "connect … CKAN" is exactly what users say) and not 5 (no synonym/extension-level coverage).

Distinctiveness Conflict Risk

The niche is narrow — connecting an existing PortalJS portal to a live CKAN instance — and the trigger explicitly distinguishes it from sibling skills ("instead of a static manifest", "connecting an existing portal" vs. scaffolding, adding datasets, or deploying). Conflict risk with other skills is minimal, matching anchor 5; anchor 4 would require noticeable overlap with closely related skills, but the related portaljs skills each have clearly separate triggers.