14 tools with consistent naming (verb_noun pattern: get_*, list_*) and complete input schemas. All tools have descriptions (avg 95 chars, within 10-1024 range). However, descriptions are generic and lack LLM-optimization guidance (e.g., 'when to use this vs similar tools', dependency hints, prerequisites). Parameters have types and descriptions, but lack constraints (enums, ranges, patterns). No output schemas documented. Error handling not visible in tool definitions. All tools are READ_ONLY, reducing security risk but limiting utility.

• Document output schemas for all 14 tools. Include field names, types, and descriptions. Example: get_document_metadata should return {title: string, word_count: integer, element_counts: {headings: integer, paragraphs: integer, ...}}.
• Enhance descriptions with LLM-optimization guidance. For get_document_text vs get_document_content, add: 'Use get_document_text for plain text extraction; use get_document_content for structured analysis with headings and formatting preserved.'
• Add enum constraints to value_format parameters. Change 'formatted (default, display values), typed (with type info), raw (underlying values as strings)' to a formal enum: ['formatted', 'typed', 'raw'].
• Document mutually exclusive parameter groups formally. For get_sheet_data, add a note: 'Exactly one of sheet_index, sheet_name, or sheet_gid must be provided. If range includes sheet name, these are ignored.'
• Add error handling guidance to descriptions. Example: 'Returns error if document_id is invalid, malformed URL, or user lacks read permission. Try search_documents() if ID is unknown.'

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

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