Server defines 8 tools with complete input schemas and basic descriptions. All tools are READ_ONLY and follow a consistent naming pattern (verb_noun). However, descriptions are terse (10-80 chars, well below the 194-char baseline), parameter descriptions are sparse or missing, and output schemas are entirely undocumented. No error handling guidance, no parameter constraints (enums), and no indication of what fields agents should expect in responses. The implementation is functional but lacks LLM-optimization for tool selection and parameter reasoning.

• Document output schemas for all 8 tools. Specify which fields are returned (e.g., search returns 'result' array with 'id', 'name', 'rank', 'status'; taxon returns 'id', 'name', 'rank', 'classification'). Include example responses.
• Expand tool descriptions to 100-200 chars. Add WHEN to use each tool: 'Use search for fuzzy name matching across ~2.2M species. Use name_match for exact scientific name lookup with optional authorship disambiguation. Use taxon to fetch full details by ID.'
• Add descriptions to all parameters. For 'id': 'Taxon ID (e.g., "1001") or name-usage ID (e.g., "1001-1") from search/name_match results.' For 'dataset': 'ChecklistBank dataset key (default "3LR" = Catalogue of Life latest release; see https://www.checklistbank.org/datasets for alternatives).'
• Add enum constraints to 'rank' and 'status' parameters in search tool. Define valid ranks (kingdom, phylum, class, order, family, genus, species, subspecies) and statuses (accepted, synonym, bare_name, misapplied, provisionally_accepted, doubtful) as JSON Schema enums.
• Enhance error messages with recovery guidance. E.g., 'Taxon ID "xyz" not found in dataset 3LR. Try search(query="xyz") to find the correct ID, or check dataset parameter.' Return available dataset keys when dataset lookup fails.