Server has 7 well-named tools with complete input schemas and reasonable descriptions (avg 85 chars). All tools follow verb_noun naming (scopus_search, scopus_get_abstract, etc.). Schemas include proper types and enums. However, descriptions lack depth, most are 1-2 sentences without WHEN/WHY guidance, dependency hints, or recovery paths. Parameter descriptions are present but minimal (avg 45 chars vs baseline 72). No output schemas documented. Error handling is absent from tool definitions. No tool annotations (readOnlyHint, etc.). All tools are READ_ONLY but this is not formally declared.

• Add output schemas to all 7 tools. Document the structure of search results (e.g., list of papers with fields: scopus_id, title, authors, year, citation_count). Use JSON Schema format in tool metadata.
• Expand tool descriptions to 150 - 250 chars. Add WHEN/WHY guidance: 'Use this to fetch full abstract and author list after scopus_search returns a paper ID. Returns null if identifier not found, try alternative identifier types (doi, eid, pubmed_id).'
• Add readOnlyHint: true to all 7 tools in schema annotations. This formally declares safety and lets LLMs reason about retry safety and composition.
• Enhance parameter descriptions with format constraints and dependency hints. E.g., 'Scopus Boolean query using TITLE-ABS-KEY(), AUTHLASTNAME(), SUBJAREA(). See scopus://search-syntax resource for syntax guide. If you only have a DOI, use scopus_get_abstract instead.'
• Add error handling guidance to tool descriptions. E.g., 'Returns empty list if no matches found, try broadening the query or using scopus_search_authors to verify author name spelling.'

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

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