Conciseness

The body is dense and entirely repo-specific — no space is spent explaining concepts Claude already knows (Playwright, OIDC, page-object pattern are used, not taught). However, there are minor instances of editorial padding that could be trimmed: "A suffix and a non-zero byte count do not prove... period", "That guard is not decoration", "The place journeys silently rot", and the multi-line block quotes of shipped docstrings restate rules already stated around them. This fits the 4 anchor ("efficient; minor instances of over-explanation that could be trimmed"); not 5 because those rhetorical passages don't each earn their tokens, and not 3 because none of it is generic background knowledge.

Actionability

Fully executable guidance throughout: the sign-in snippet ("from .flows import sign_in / sign_in(page, base_url, session_for(\"dev_ledit\"))"), concrete locator examples ("page.get_by_role(\"dialog\", name=\"Git output\")"), a copy-paste manifest-derivation snippet with its guard, and exact run commands with gotcha callouts ("uv run --project tests --frozen playwright install chromium", "./dev-compose.sh test-stand test -k login --headed"). File-level responsibilities are pinned in a table. This is the 5 anchor — copy-paste-ready code and commands covering the common cases.

Workflow Clarity

The 10-step Procedure is a clear sequence with explicit validation checkpoints and feedback loops: step 1 gates on "Try to make it an API test first", step 2 requires a measured justification, step 9 requires "run headed once to confirm it fails for the right reason when you break the expectation deliberately", and "Run `stand-test-auditor` over a new journey before considering it done" supplies an external audit loop. Not 4: validation is not merely mentioned but given as a deliberate red-green check (break the expectation, watch it fail), which is the 5 anchor's "explicit validation steps; feedback loops for error recovery".

Progressive Disclosure

No bundle files exist (no references/, scripts/, or assets/), so everything — the download/normalizer contract details, the locator-exception catalog, quoted exemplar docstrings, and run-mode notes — is inlined in one ~280-line SKILL.md. Section headers make it navigable and sibling-skill references (insight-stand, stand-scenarios, stand-test-auditor, quality-vector-tests) are clearly named, but content that belongs in reference files is inline, which is the 3 anchor ("some structure... content that should be separate is inline"). Not 4: at this length, splitting the worked-example quotes and the cross-format download comparison into one-level-deep reference files would materially reduce the always-loaded surface; not 2 because the file is well-sectioned and easily navigated, not a headerless dump.

Specificity

Names multiple concrete actions — "Write, fix, or review the browser journeys in tests/stand/ui/ — Playwright (Python, sync API) against a deployed Insight stand with a real Keycloak sign-in" — and comprehensively enumerates capability areas: "the governing UI-vs-API rule... accessibility-first locators, the page-object/flows split, manifest-derived expectations, URL/deep-link validation, and real browser downloads". This matches the 5 anchor ("multiple specific concrete actions; comprehensive coverage") with no coverage gaps; a 4 would require a missing capability area, and none is evident.

Completeness

Explicitly answers both questions: "what" in the first two sentences (write/fix/review browser journeys plus the enumerated coverage areas) and "when" via "Use for any request to write, add, fix or review a committed browser test in this repository" with concrete trigger phrases. This is exactly the 5 anchor pattern ("clearly and explicitly answers both what AND when with concrete trigger phrases"); not 4, because the when-clause is as explicit and specific as the anchor example.

Trigger Term Quality

Comprehensive natural trigger phrases with synonyms and path forms: "'add a Playwright test for X'", "'the UI test is failing'", "'turn this scenario into a browser journey'", "'add a deep-link or download regression'", "anything under tests/stand/ui/". This is the 5 anchor ("comprehensive coverage of natural terms including synonyms and file extensions") — the quoted phrases are exactly what a user would type; a 4 would mean a common natural variant is missing, and none is.

Distinctiveness Conflict Risk

Clear niche (committed browser tests under tests/stand/ui/) with explicit boundaries against every adjacent skill: "For HTTP contract tests use stand-api-test; playwright-cli is for driving a browser interactively at a prompt... reach for it only when nothing will be committed under tests/stand/ui/; drive-ui is for looking at a stand by hand". Minimal conflict risk — the 5 anchor; not 4, because the disambiguation goes beyond minor overlap avoidance and actively routes each neighboring use case elsewhere.