Server has 5 tools with complete input schemas and basic descriptions. Naming follows verb_noun convention (shorten_url, get_url_info, delete_url, list_urls). However, descriptions are generic and lack LLM-optimized guidance on when to use each tool vs. alternatives. Parameter descriptions are minimal (e.g., 'The short code to look up' lacks format/length constraints). Output schemas are not documented, LLMs cannot predict response structure. Error handling returns plain text errors without recovery guidance. No tool annotations (readOnlyHint, destructiveHint) despite clear risk levels (WRITE, DESTRUCTIVE). Missing idempotency guidance for retry-safe operations.

• Add output schema documentation to each tool. Example: shorten_url returns {code: string, shortUrl: string, qrCode: string, createdAt: ISO8601}. This enables LLMs to chain tools and extract correct fields.
• Enhance parameter descriptions with format/constraint details. E.g., 'alias: Optional custom alias (2-20 alphanumeric chars, must be unique)' and 'expiresIn: Optional expiration time in seconds (60-31536000, default: never expires)'.
• Add tool annotations to inputSchema. Mark delete_url with destructiveHint: true, mark get_url_info/get_url_stats with readOnlyHint: true, mark shorten_url with idempotentHint: false (if alias is provided, it's idempotent; otherwise non-deterministic).
• Improve error messages with recovery guidance. E.g., 'URL not found. Verify the short code is correct, or call list_urls() to see available URLs.' Categorize errors: retryable (timeout, rate limit) vs. user-fixable (not found, invalid input).
• Clarify tool selection in descriptions. E.g., get_url_info: 'Retrieve metadata (original URL, creation date, click count) for a shortened URL.' get_url_stats: 'Retrieve detailed click analytics (referrer, user agent, timestamp) for a shortened URL. Use this for reporting; use get_url_info for quick metadata.'