Four tools with complete input schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. Tool names follow verb_noun convention (azure_*). Descriptions range 80 - 250 chars, meeting minimum length. However, most parameters lack type constraints (enums, min/max), and output schemas are not documented. No error handling guidance visible. Parameters like 'category', 'impact', 'search' in azure_advisor_recommendations lack enum constraints despite accepting known values. azure_public_exposure_inventory has empty required array and no parameters documented. Composition is reasonable (each tool has one responsibility), but parameter relationships and dependencies are underdocumented.

• Add enum constraints to azure_advisor_recommendations: category=['cost','performance','security','reliability','operational_excellence'], impact=['high','medium','low'].
• Document output schema for each tool. Example for azure_advisor_recommendations: {recommendations: [{id, title, description, category, impact, savings_estimate}], total_count, next_cursor}.
• Add enum constraint to run_performance_profile window parameter: window=['PT1H','PT6H','P1D','P7D','P30D'].
• Expand azure_public_exposure_inventory description to explain what 'public endpoints' means and when to call it (e.g., 'Returns all Azure resources with public IP addresses or internet-facing endpoints. Call this first for security audits.').
• Add recovery hints to descriptions: 'If subscription not found, call azure_resource_inventory with request="subscriptions" to list available subscriptions.'