Three tools with clear names and reasonable descriptions, but significant gaps in schema completeness and parameter documentation. All tools lack output schema documentation, and several parameters are missing type constraints. The 'query' tool has complex SoQL parameters that need better validation guidance. No error handling patterns visible. Descriptions are adequate (100-200 chars) but could be more prescriptive about when to use each tool.

• Document output schemas for all three tools. E.g., 'datasets returns {results: [{resource_id, name, description, category, updated_at}], total_count}'. This enables LLMs to chain calls and extract IDs.
• Add min/max constraints to numeric parameters: limit 1-100 (default 20), offset >= 0. For 'query' tool, add limit default 1000 and document Socrata's max row limit.
• Replace example values in descriptions with formal constraints. Instead of 'e.g. "budget"', use 'keyword string (e.g., budget, crime, health)' or better yet, move examples to a separate 'examples' field if the schema supports it.
• Add error handling with recovery hints. E.g., if resource_id is invalid, return 'Dataset not found. Call datasets() to search for valid resource IDs.' If SoQL syntax fails, return the Socrata error and suggest 'Check SoQL syntax: valid clauses are where, select, group, order.'
• Document the relationship between tools: 'Workflow: (1) datasets() to find resource_id, (2) metadata() to inspect schema, (3) query() to fetch data.' This guides LLM planning.