Two tools with basic schemas and minimal descriptions. Both tools have input schemas with proper JSON Schema structure and type definitions. However, descriptions are extremely brief (10-27 chars), falling well below the 50-200 char baseline for LLM-optimized descriptions. Parameter descriptions are present but terse. No output schema documentation. No error handling guidance. No tool annotations (readOnlyHint, etc.). Tool names lack action verbs ('hello' and 'add' are generic; 'greet_user' and 'sum_numbers' would be clearer). No pagination, no composition hints, no recovery guidance.

• Expand 'hello' description to: 'Greet a person by name. Returns a friendly greeting message. Use this to acknowledge users or test the server. If name is omitted, defaults to "World".'
• Expand 'add' description to: 'Add two numbers and return their sum. Both parameters are required. Returns the result as a formatted string. Use for arithmetic operations.'
• Rename 'hello' → 'greet_user' and 'add' → 'sum_numbers' to follow verb_noun convention and improve LLM intent inference.
• Document output schema for both tools: specify that responses contain a 'content' array with 'type' and 'text' fields.
• Add tool annotations: both tools are read-only and idempotent; mark with readOnlyHint=true and idempotentHint=true in the tool definition.

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

• Relies on Stateful initialize / Mcp-Session-Id (removed; protocol is stateless) - make each request self-contained