Server has 7 well-named tools with mostly complete schemas and descriptions. Naming follows verb_noun convention (list_, set_, get_, fetch_, search_, refresh_). Descriptions are present and substantive (avg ~150 chars), explaining WHAT and WHEN. Input schemas are properly typed with enums where appropriate (render, format, sort). However, output schemas are not documented, LLMs cannot predict response structure. Error handling returns structured {error, message} but lacks recovery guidance. Tool annotations (read_only_hint) are correctly applied to 5 read-only tools. Missing: pagination guidance for search/list results, parameter constraints (e.g., max_chars 500-200000 stated but not enforced in schema), and explicit output field documentation.

• Document output schemas for all tools. E.g., 'Returns {resources: [{name, slug, target, enabled, search_adapter}], total: N}' for list_resources. Use JSON Schema or explicit field descriptions.
• Add min/max constraints to numeric parameters in schema. fetch_url.max_chars should declare minimum: 500, maximum: 200000 in the JSON Schema, not just the description.
• Enhance error messages with recovery hints. E.g., if set_resource_enabled fails with 'Resource not found', append 'Call list_resources() to see available databases.'
• Add pagination to search() results. Include offset/limit in request, and return {results: [...], total: N, has_more: bool} so agents know when to fetch next page.
• Clarify get_download_url output. Document: 'Returns {url: string, title: string, authors: [string], year: int, doi: string} or {error, message} if no PDF available.'

Inferred effective spec: <=2025-11-25.

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