Conciseness

The body is mostly command documentation without concept over-explaining, but contains several instances of padding: the URL-normalization explanation ('The URL is normalized to a domain and matched against known apps. If no app is found, one is created and a connector is built automatically'), a duplicated action-discovery section ('Searching for actions' and 'Popular actions' both show `action list --intent`), and a garbled filler sentence ('Those will be used to adjust tooling to be used best with your harness'). It sits between anchors 2 and 3, closer to 3, since the majority of lines are lean executable commands.

Actionability

Concrete, executable commands throughout — `npm install -g @membranehq/cli@latest`, `membrane connection ensure "https://checkr.com" --json`, `membrane action run <actionId> --connectionId=... --input '{...}'` plus a complete proxy flag table. The gap keeping it from 5: no worked Checkr-specific example at all — the 'Popular actions' section contains no actual actions (no create-candidate, run-background-check, or invitation examples), so common cases are not covered.

Workflow Clarity

A clear install → login → connect → discover → run sequence with genuine validation loops (poll `BUILDING` until ready, handle `clientAction` states, check `error` field on failure). It falls short of 5 because the step references are incoherent: 'skip to **Step 2**' points to a step that doesn't exist, and '#### 1b. Wait for the connection' appears with no 1a — minor validation gaps and broken cross-references, matching anchor 4 rather than 5.

Progressive Disclosure

No bundle files exist (references/, scripts/, assets/ are absent), so all content lives in a single ~140-line SKILL.md with clear section headers and an official-docs pointer. Structure is good and most content is appropriately placed; the inlined generic Membrane CLI lifecycle material (connection states, proxy flags) and the empty 'Popular actions' and malformed 'Checkr Overview' entity lists are minor organization gaps, matching anchor 4 rather than 3 (structure is clearly good) or 5 (over 50 lines with content that could be split out).

Specificity

'Manage data, records, and automate workflows' names the domain ('Checkr integration') but the actions are generic to the point of near-meaninglessness — no background checks, candidates, invitations, or packages are mentioned. This matches anchor 2 ('Names the domain but actions are minimal or generic') and is below anchor 3, which requires 1-2 concrete actions.

Completeness

Both parts are present: what ('Checkr integration. Manage data, records, and automate workflows') and an explicit when ('Use when the user wants to interact with Checkr data'). Structurally this matches anchor 4 ('Has both what and when; when could be more explicit or specific') rather than 3, since the 'when' is explicitly stated, not merely implied — though both halves are generic, and the 'what' is weaker than anchor 4's example.

Trigger Term Quality

'Checkr' and 'records' are relevant keywords, but the description misses the natural phrases users would actually say — 'background check', 'background screening', 'candidate', 'vetting' — none of which appear. Fits anchor 3 ('Some relevant keywords but missing common variations or synonyms'); not 4 because the single most important synonym family ('background check') is absent.

Distinctiveness Conflict Risk

The 'Checkr' proper noun carves out a clear niche and the trigger requires Checkr, so wrong-skill triggering is unlikely. Minor overlap risk remains from the generic 'automate workflows' and 'manage data' phrasing, which could collide with workflow-automation or data-management skills — matching anchor 4 ('Mostly distinct; minor overlap risk') rather than 5.