Severe structural issues undermine this server. Tool definitions are duplicated (search_notes, refine_search, get_stats appear 3-4 times with inconsistent schemas). Descriptions are present but minimal (10-50 chars), below the 50-200 char LLM-optimized baseline. Parameter descriptions exist but lack detail on constraints, formats, or valid ranges. No output schemas documented. Critical: tool definitions appear scattered across legacy files (Cloudflare Workers, Flask, FastMCP) with no unified registration visible in the main source. The fastmcp framework code is incomplete in the provided snippet, making it impossible to verify actual tool registration.

• Consolidate duplicate tool definitions into a single, canonical registration. Remove variants with/without api_key, use server-side environment variable injection (API_KEY env var) instead of tool parameters.
• Expand all descriptions to 50-200 chars. Example: 'search_notes' → 'Search Apple Notes using semantic embeddings. Returns up to 20 results ranked by relevance. Supports natural language queries (e.g., "funny jokes", "work notes"). Use refine_search for date filtering.'
• Document output schemas for each tool. E.g., search_notes returns [{title, content, updated, score}]. get_stats returns {indexed_notes, model, dimensions, status}. Include field types and meanings.
• Add parameter constraints: limit (1-20, default 5), query (1-500 chars), date_after/date_before (ISO 8601 format, optional). State these in descriptions and JSON Schema.
• Add error handling guidance. E.g., 'If no results found, try a shorter query or remove date filters.' For refresh_index, document what happens if indexing fails and how to retry.