Two tools with minimal descriptions and incomplete schemas. list_tools has empty input schema (correct for no-param tool) but description is generic. call_tool has a basic schema with two parameters but lacks critical details: no enum constraints on tool names, no description of what 'arguments' object structure should be, no guidance on error cases. Neither tool documents output structure. Descriptions are under 200 chars but lack actionable context for LLM selection. No error handling guidance, no parameter validation rules, no security warnings about routing untrusted calls.

• Add detailed description to call_tool parameter 'arguments': 'Tool arguments as a JSON object. Structure depends on the target tool schema. Validate against the tool's inputSchema from list_tools before calling.'
• Document call_tool output schema: specify that it returns content array with TextContent/ImageContent/EmbeddedResource objects, and include example structure.
• Add enum constraint or validation guidance for call_tool 'name' parameter: 'Must be a valid tool name from list_tools(). Invalid names return error: "Tool not found".'
• Expand list_tools description to explain filtering behavior: 'Returns tools available to this agent based on gateway configuration. Some tools may be blocked, renamed, or virtualized. Call this first to discover available tools.'
• Add error handling guidance to call_tool: 'If backend server is unreachable, returns error with server name. If tool not found, returns error with available tool names. If arguments invalid, returns schema validation error.'

Inferred effective spec: 2026-07-28+.

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