Conciseness

The body is lean and assumes competence throughout — it explains only instance-specific behavior (the <instance-context> block, the 20-entry expand cap, retention windows) and nothing Claude would already know. Every section teaches something non-obvious about this instance's surfaces.

Actionability

Concrete, copy-paste-ready tool calls with exact parameters are given for rungs 1–2 (`activity(action="expand", id=N)`, `workflows(action="get", workflowId, full=true)` with the `full=true` flag explained) and for credentials. The gap: `executions` is referenced three times ("`executions` has the detail") but its call syntax is never shown, which keeps this just below anchor 5.

Workflow Clarity

A clearly sequenced escalation ladder ("Read in this order — cheapest first", "Stop at the first one that answers yours") with explicit stop conditions, a question-to-surface decision table, and error-recovery guidance ("An id that no longer resolves is ordinary — carry on rather than treating it as an error"). This is a read-only lookup skill, so the destructive/batch validation cap does not apply.

Progressive Disclosure

Well-organized single-file structure with clear sections, a lookup table, and no nested or dangling references — and all content genuinely belongs inline since it is operational guidance rather than reference material. It scores 4 rather than 5 because the body is ~108 lines, beyond the under-50-line simple-skill exception, and a couple of sections (e.g. the per-surface detail) could live in a reference file if the skill grows.

Specificity

The domain is clearly named ("what is already on this instance", "what has recently run or broken") but the only action verb is "Load" — no concrete capabilities (e.g. looking up activity, inspecting failed runs, reading workflows) are explicitly stated. This matches anchor 3 (domain named, capability areas hinted but not comprehensively listed) rather than 4, which requires specific listed actions.

Completeness

The "when" is explicit and rich, but the "what" is only weakly implied — the description never states what the skill actually does (look up instance state via activity/workflows/executions), only when to load it. Anchor 3 (one side clear, the other only implied) fits; it is not 4 because the what is never explicitly stated, and not 2 because the domain does come through.

Trigger Term Quality

Quotes natural openers users would actually say ("fix it", "carry on", "what should I look at") and adds paraphrased variants ("a reference to earlier work in a new conversation", "any question about what has recently run or broken"), giving comprehensive natural-term coverage for this niche.

Distinctiveness Conflict Risk

The instance-state scope ("what is already on this instance rather than on what the user just typed") carves a distinct niche, but very broad triggers like "a short or ambiguous opener" could plausibly fire other context/memory skills — minor overlap with closely related skills, matching anchor 4 rather than the minimal-conflict profile of 5.