Three tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. Tool names follow verb_noun pattern (search_, crawl_, verify_) which is good. Descriptions are present but generic (10-50 chars for most params). No parameter descriptions for critical fields like 'urls' array items or 'pages' object structure. Output schemas are not documented, LLMs cannot predict response structure. Error handling is minimal. No tool annotations (readOnlyHint present in metadata but not in schema). Chinese descriptions reduce LLM clarity.

• Add explicit output schema documentation for each tool. E.g., search_company returns {success: bool, results: [{title: string, url: string, snippet: string}], count: int}. This enables LLM to chain tools and extract fields correctly.
• Add descriptions to all parameters. For 'urls' in crawl_multiple_pages: 'Array of LinkedIn profile or company page URLs to crawl in parallel. Max 10 URLs per request. Must be valid HTTPS URLs.'
• Translate all descriptions to English. Replace Chinese descriptions with English equivalents for maximum LLM compatibility and clarity.
• Add input validation constraints to schemas. E.g., company_name: minLength 1, maxLength 100; urls: minItems 1, maxItems 10; pages: minItems 1, maxItems 50.
• Enhance error responses with recovery guidance. E.g., 'Search failed: API rate limit exceeded. Retry in 60 seconds or try a shorter company name.' instead of bare error JSON.