Server defines 8 tools with consistent verb-noun naming (search_*, download_*). All tools have descriptions (10-60 chars, mostly adequate). Input schemas are present with type definitions and parameter descriptions. However, output schemas are not documented, LLMs cannot predict response structure. Error handling is minimal; no recovery guidance or actionable error messages. Parameter descriptions lack format constraints (e.g., date_from/date_to accept strings but no validation hints). No tool annotations (readOnlyHint/destructiveHint) despite clear read/write distinction. Missing pagination guidance for search tools that return lists.

• Document output schemas for all search tools. Specify: array of objects with fields (paper_id, title, authors, date, url, abstract). Include total_count and next_cursor for pagination.
• Add tool annotations: readOnlyHint=true for all search_* tools; destructiveHint=true for all download_* tools. Enables agent safety checks.
• Add format constraints to date/year parameters. E.g., 'date_from: ISO 8601 string (YYYY-MM-DD), e.g., 2024-01-01' and 'year_min: integer 1900 - 2100'.
• Implement pagination for search results. Add offset/limit or cursor parameters. Document max_results cap (e.g., 'max 100 results per call; use pagination for more').
• Enhance error messages with recovery guidance. E.g., 'Paper ID invalid. Verify with search_arxiv_papers first.' Return structured error objects with code, message, and suggested_action.