Three tools with minimal definitions. All lack verb-prefixed names (applications, sites, users are nouns, not actions). Descriptions are generic and under 100 chars, providing no context on WHEN to use each tool or WHAT distinguishes them. Input schemas are present but trivial: single optional 'name' string with no type validation, enums, or constraints. No output schema documented. No error handling guidance. No parameter relationships or dependencies explained. Tools appear to be simple read-only wrappers around Microsoft Graph but lack the specificity and LLM-optimization required for reliable agent selection.

• Rename tools with action verbs: 'list_applications' (all), 'search_applications' (by name), 'get_application' (by ID). Clarify intent in the name itself.
• Expand descriptions to 150-250 chars. Example: 'List all Azure AD applications or search by display name. Returns application ID, display name, app ID, publisher domain, and creation date. Use this to discover available apps before querying their properties.'
• Document output schema explicitly. Specify that response is a JSON object with application IDs as keys and nested objects containing {id, displayName, appId, publisherDomain, createdDateTime}.
• Add constraints to 'name' parameter: 'Optional. Display name of the application to filter by (case-sensitive substring match). If omitted, returns all applications (paginated, max 20 per call).'
• Add error handling: return structured errors like '{"error": "not_found", "message": "No applications found matching name: xyz", "suggestion": "Try a partial name or omit the filter to list all."}' instead of generic text.