Server provides 6 well-named, read-only tools with complete input schemas and descriptions. All tools follow verb_noun naming (search_*, get_*, list_*) and include parameter descriptions. However, output schemas are not explicitly documented in the source code, only input schemas are visible. Parameter descriptions are present but brief (avg ~40 chars), lacking actionable constraints like ranges, formats, or enums. Tool descriptions are adequate (avg ~50 chars) but could better explain WHEN to use each tool vs. alternatives. No error handling guidance visible in tool definitions. Security is strong (read-only, no secrets), but composition could be clearer, e.g., search_items vs. search_books distinction is not well explained.

• Document output schemas for all 6 tools in src/server.ts or tool definition files. Include field names, types, and descriptions. Example: search_items returns {items: [{identifier, title, creator, date}], total_count, page}.
• Add min/max constraints to numeric parameters: limit (1 - 100), page (1+). Add format constraint to date parameter: 'YYYY-MM-DD format, e.g. 2024-01-15'.
• Expand tool descriptions to 100 - 150 chars, explicitly stating WHEN to use each tool. Example: 'Search the Internet Archive catalogue by title, creator, or keyword. Use this to find items; use search_inside to search text within a specific item.'
• Add error handling guidance to descriptions: 'Returns empty results if no matches found. If a specific item is not found, try search_items with a partial title.'
• Standardize pagination: use cursor-based pagination across all search tools, return total_count and next_cursor in output schema, and document in tool descriptions.

Inferred effective spec: 2025-06-18+.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry