Three well-named, read-only tools with clear descriptions and complete input schemas. All tools follow verb_noun naming (search_*, query_*, layer_info). Descriptions are concise (100 - 200 chars) and explain WHAT, WHEN, and WHAT-TO-DO-NEXT. Input schemas include types and descriptions for all parameters. However, output schemas are not documented, LLMs cannot predict response structure. Parameter descriptions lack format constraints (e.g., 'limit: 1 - 50' is stated but not formalized). No error handling guidance or recovery paths. SSRF guard in layerUrl() is a security plus, but not reflected in tool descriptions.

• Document output schemas for all three tools. Example for search_datasets: 'Returns {total: number, count: number, datasets: [{id, name, summary, record_count, url}]}'. This enables agents to plan downstream calls.
• Add error handling guidance. Replace 'throw new Error("ArcGIS Hub: 404")' with 'throw new Error("Dataset not found. Try a broader keyword or check org_id. Available orgs: ...")'. Guide LLMs on retry vs. user input.
• Formalize parameter constraints in JSON Schema. Add minValue/maxValue to limit params: {limit: {type: 'number', minimum: 1, maximum: 50, description: '...'}}. This prevents invalid values at the schema level.
• Clarify URL validation in tool descriptions. Add: 'url must be a public ArcGIS Feature/Map Service endpoint (https, no IP addresses or localhost). Example: https://services.arcgis.com/...FeatureServer/0'.
• Reduce response verbosity. Strip org_id, type, and other metadata from search_datasets response. Return only: {total, datasets: [{id, name, summary, record_count, url}]}. Saves tokens and reduces noise.