Server has 15 tools with consistent naming (verb_noun pattern: search_, query_, get_, list_) and descriptions present for all. However, descriptions are often terse (10-50 chars), parameter descriptions lack detail on constraints/formats, and output schemas are not documented in the visible code. Tools are read-only and well-scoped (one responsibility each), but lack error recovery guidance and parameter validation hints. No tool annotations (readOnlyHint, etc.) despite all being read-only. Schemas visible for ping and inferred from evaluation scripts for others, but not all explicitly registered with full input/output schema definitions in the main server code.

• Add explicit output schemas to all tools. Document returned fields, types, and structure. Example: socrata_search_datasets should declare it returns {datasets: [{id, name, description, columns}], total_count}.
• Expand parameter descriptions with constraints. Add min/max for 'limit' (e.g. '1-100, default 20'), enum values for 'lang' ('ca|es|en'), and format hints for 'query' ('free-text search string, 2-200 chars').
• Extend tool descriptions to 100-150 chars. Include when to call (e.g. 'Call this first to discover available datasets'), what it reveals (structure/metadata), and dependencies (e.g. 'Requires a dataset ID from socrata_search_datasets').
• Add readOnlyHint: true to all tool definitions via tool annotations. This signals to agents that calls are safe to retry and have no side effects.
• Document error scenarios in descriptions. E.g. 'Returns error if dataset_id not found; try socrata_search_datasets to find valid IDs.' This enables agent self-correction.

Inferred effective spec: 2025-06-18+.

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