Server defines 5 tools with complete schemas and descriptions. All tools are READ_ONLY and follow verb_noun naming (query_, get_, list_). Descriptions are present (avg 140 chars) and explain what each tool does. However, parameter descriptions are minimal or absent for several tools (e.g., get_contradictions and get_graph_stats have empty properties with no descriptions). Output schemas are not documented, callers cannot predict response structure. Error handling is basic (returns text responses without recovery guidance). No tool annotations (readOnlyHint, idempotentHint) despite all being read-only. Missing pagination support on list_entities despite potentially returning large result sets.

• Document output schemas for all tools. E.g., query_knowledge_graph should return {answer: string, sources: [{doc: string, claim: string}]}. This enables LLMs to extract and chain results.
• Add readOnlyHint: true to all 5 tool definitions in list_tools(). This signals to agents that these tools are safe to call without side effects.
• Add pagination to list_entities: accept optional limit (default 50, max 500) and offset parameters. Return {entities: [...], total: N, offset: M} so agents can iterate large result sets.
• Enhance error messages with recovery hints. E.g., 'No claims found for entity "XYZ". Call list_entities to see available entities, or try a partial name match.'
• Add parameter descriptions to get_contradictions and get_graph_stats inputSchema, even if empty (e.g., 'This tool takes no parameters'). Explicit is better than implicit.