Conciseness

At ~27 lines the body is lean and assumes competence: no background on what Azure Advisor is, and every line instructs ('Prefer Azure Advisor MCP tools over Resource Graph or generic ARM tools', 'Never count a capped list', 'Resource-group scope requires a subscription'). Nothing is padded and nothing could be trimmed without losing guidance.

Actionability

Every intent maps to a named MCP tool — advisor_recommendation_summary, advisor_recommendation_list, advisor_recommendation_update, advisor_metadata_list/get, advisor_remediation_get — including the exact state-changing verbs ('Complete, dismiss, postpone, or reactivate') and scope rules. As an instruction-only skill the guidance is fully actionable and covers the common cases.

Workflow Clarity

The four-step Workflow is clearly sequenced and includes checkpoints for the risky operations: 'Confirm before recommendation-state updates' and 'Present only returned data, including truncation, access limitations, and errors'. Not 5 because there is no explicit error-recovery feedback loop (e.g., what to do when an advisor_recommendation_update call fails); not 3 because the state-update operation does have an explicit confirmation gate, so the missing-validation cap does not apply.

Progressive Disclosure

The skill is under 50 lines with no bundle files, and the body is cleanly organized into Quick reference / Workflow / Safety sections with the single cross-skill pointer to azure-advisor-actionability clearly signaled ('Retrieve remediation → advisor_remediation_get, then azure-advisor-actionability'). This matches the simple-skill exception: well-organized sections alone warrant 5.

Specificity

The declarative 'what' is generic — 'Use Azure Advisor recommendations' names the domain with a single verb — and the only explicitly concrete action, 'update recommendation status', appears inside the WHEN trigger quotes rather than as a stated capability. Not 2 because a concrete action is explicitly stated beyond domain-naming; not 4 because the skill's other capabilities (summarize, list, browse metadata) are never stated as actions.

Completeness

Both parts are present: a 'what' ('Use Azure Advisor recommendations') and an explicit, concrete WHEN clause, so the missing-'Use when' cap at 3 does not apply. Not 5 because the 'what' is vague — it never states what the skill does with recommendations (view, aggregate, update status); not 3 because the 'when' is explicit with concrete trigger phrases.

Trigger Term Quality

The WHEN clause quotes natural user utterances — "show Advisor recommendations", "service retirements", "update recommendation status" — that a user would plausibly say. Not 5 because common variations and synonyms are missing (e.g., 'dismiss', 'postpone', 'Advisor score', 'Azure recommendations'); not 3 because these are specific natural phrases rather than generic or purely technical keywords.

Distinctiveness Conflict Risk

The description names a specific Azure service and adds an explicit exclusion boundary — 'DO NOT USE FOR: fixes or remediation (use azure-advisor-actionability)' — giving it a clear niche with distinct triggers and minimal conflict risk, matching the anchor-5 example structure.