Server exposes 2 tools with minimal documentation. Both tools have trivial descriptions (10-30 chars), no parameter descriptions despite accepting inputs, and no documented output schemas. Tool naming follows verb_noun convention (getAllAccounts, getAccountDetailsById) but descriptions lack context for LLM selection. No error handling guidance, no pagination support despite list operations, and no security annotations. This is a skeleton implementation missing nearly all production-grade patterns.

• Expand tool descriptions to 50-200 chars. Example: 'getAllAccounts: Retrieve a paginated list of all bank accounts accessible to the authenticated user. Returns account ID, type, balance, and status. Use this to discover available accounts before querying details.'
• Add parameter descriptions. For accountId: 'The unique account identifier (format: ACC-XXXXX or 12-digit number). Obtain from getAllAccounts() or your account statement.'
• Document output schemas explicitly. Specify that getAllAccounts returns {accounts: [{id, type, balance, status, currency}], total_count, next_cursor} and getAccountDetailsById returns {id, type, balance, status, currency, owner, created_at, last_transaction}.
• Add pagination to getAllAccounts: accept limit (1-100, default 20) and offset/cursor parameters. Return total_count and next_cursor to enable iteration over large account lists.
• Add error handling with recovery hints. E.g., 'Account ACC-99999 not found (404). Available accounts: ACC-12345, ACC-67890. Call getAllAccounts() to see all accounts.'

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

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