EduAssist has 5 tools with basic schemas and descriptions, but significant gaps prevent production use. All tools have descriptions (10-100 chars, mostly adequate), and input schemas are present with types. However, parameter descriptions are sparse or missing context. No output schemas documented. Error handling is absent, tools provide no recovery guidance. Tool names follow verb_noun convention (find_, search_, extract_, get_, fetch_) but lack specificity about when to use each. The fetch tool is a pass-through to an external MCP server, not a native implementation. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite all being read-only. Composition is weak: search_papers and extract_info both operate on papers but with unclear distinction. No pagination support visible despite search_papers returning multiple results.

• Document output schema for each tool. Example: search_papers returns [{paper_id: string, title: string, authors: [string], url: string, abstract: string, published_date: string}]. Include all fields the agent needs for downstream calls.
• Add parameter format guidance. For extract_info, clarify: 'paper_id: arXiv identifier (e.g., 2301.12345 or 2301.12345v2)'. For search_papers, specify: 'topic: free-form search query (e.g., "machine learning", "neural networks")'. For get_transcript, state: 'url: full YouTube URL (https://www.youtube.com/watch?v=...).'
• Add tool annotations. Mark all tools with readOnlyHint: true since they do not modify state. This signals to agents that these calls are safe to retry.
• Implement error recovery guidance. Example: 'If paper not found, try search_papers() with a broader topic. If transcript unavailable, suggest watching the video directly or searching for a transcript on the web.'
• Clarify tool composition. Rename extract_info to extract_paper_info and add description: 'Search for detailed information about a specific paper (by arXiv ID) across cached directories. Use after search_papers() to get full details.' This disambiguates from search_papers.