Three tools with basic schemas and descriptions, but significant gaps in parameter documentation and output structure. Tool names follow verb_noun convention (add_term, list_terms, list_terms_topics). Descriptions exist but are minimal (10-50 chars) and lack actionable context for LLM selection. Parameters have types but lack detailed constraints, ranges, or format guidance. No documented output schemas. Error handling returns plain strings without recovery guidance. Validation occurs server-side but errors are not structured for LLM self-correction.

• Expand tool descriptions to 100-200 chars. For list_terms: 'Retrieve flashcards from your TermLoop dictionary, optionally filtered by topic. Returns up to 200 cards with ID, term, definition, topic, and priority. Call this to review existing cards or check if a term is already saved.'
• Return structured JSON instead of plain strings. add_term should return {"card_id": 123, "term": "...", "topic": "...", "message": "..."} so LLMs can extract IDs for follow-up operations.
• Add explicit constraints to parameters: term (1-200 chars, no leading/trailing whitespace), definition (1-2000 chars), topic (optional, 1-50 chars, alphanumeric + spaces). Use JSON Schema minLength/maxLength and pattern fields.
• Implement pagination for list_terms: add optional 'limit' (1-100, default 20) and 'offset' (default 0) parameters. Return {"cards": [...], "total": N, "offset": O, "limit": L} so agents can iterate through large dictionaries.
• Wrap error responses in structured format: {"success": false, "error": "...", "recovery": "...", "retryable": true/false}. E.g., validation errors should suggest 'Reduce term length to ≤200 chars' rather than a bare error message.