Server has 5 well-named tools with complete input schemas and Japanese descriptions. All tools follow verb_noun naming (search_, get_, list_, compare_). Schemas include types and descriptions for all parameters. However, descriptions are in Japanese only (not English), limiting LLM accessibility in multilingual contexts. Output schemas are partially documented in docstrings but not formally declared. Error handling is minimal, tools return generic error dicts without recovery guidance. No tool annotations (readOnlyHint, idempotentHint) despite all being read-only operations. Parameter descriptions lack format constraints (e.g., section enum values are listed in text but not as formal enums in schema).

• Add English descriptions to all tools. Use 50 - 150 character descriptions that state WHAT the tool does and WHEN to use it (e.g., 'Search Japanese securities filings by keyword, company, or risk section. Use this to find relevant regulatory disclosures.').
• Convert section parameter to enum in schema: {"type": "string", "enum": ["risk_factors", "md_and_a", "business_overview", "corporate_governance", "financial_summary", "full_text"]}. Remove example values from description.
• Add readOnlyHint: true to all tool definitions via FastMCP annotations or metadata. This signals to LLMs that these tools are safe to call without side effects.
• Enhance error responses with recovery guidance. Instead of {"error": "Filing not found"}, return {"error": "Filing S100VWVY not found", "suggestion": "Try list_indexed_companies() to find valid doc_ids, or search_filings() to locate filings by company name."}.
• Document output schemas formally. Add a 'Returns' section to each docstring with JSON Schema or structured description of response fields (e.g., search_filings returns list of {doc_id, company_name, section, snippet, filing_date}).