Server has 9 tools with consistent naming (verb_noun pattern: search-, get-, check-) and all tools have descriptions. However, descriptions are generic and lack actionable context. Parameter descriptions exist but are minimal (avg ~40 chars vs baseline 72). No output schemas documented. Error handling is basic (formatError wraps exceptions as text). All tools are READ_ONLY, reducing risk but also limiting utility. Schema definitions use Zod with types, but lack constraints (enums, ranges, patterns). No tool annotations (readOnlyHint, idempotentHint) despite all being read-only.

• Document output schemas for all tools. Example for search-products: {products: [{code: string, name: string, price: string, stock: string}], pagination: {totalResults: number, currentPage: number, pageSize: number}}
• Replace example-based parameter descriptions with formal constraints. Change 'Sort order (e.g., name:asc, price:desc)' to enum: ['name:asc', 'name:desc', 'price:asc', 'price:desc']. Specify pageSize range: 1-100.
• Add tool annotations to Zod schemas: z.string().describe(...).meta({readOnly: true}) for all tools, enabling LLM awareness of retry safety.
• Enhance error messages with recovery guidance. Example: 'Product code ABC123 not found. Try search-products(query="ABC") to find similar products.'
• Clarify tool selection by adding WHEN clauses to descriptions. Example: 'Use search-products for keyword-based discovery. Use search-products-advanced when you need price filtering or category constraints.'

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

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