Two tools with basic schemas and minimal descriptions. get-daily-news has empty input schema ({}), search-news has typed parameters (keyword: string, days: number with default). Both descriptions are under 100 chars and lack context on when to use them, prerequisites, or what the output structure contains. No parameter descriptions beyond the schema. No documented output schemas. Error handling returns generic text without recovery guidance. Tool names follow verb_noun pattern (positive), but descriptions are too terse to guide LLM selection effectively.

• Add explicit input schema to get-daily-news. Even if no parameters are required, document the empty schema with a description: 'No parameters required. Fetches today's news from The Verge RSS feed.'
• Expand tool descriptions to 100-200 chars. Example for search-news: 'Search The Verge news articles by keyword. Returns up to 5 matching articles with title, link, publication date, and summary. Useful for finding coverage of specific topics or companies.'
• Document output schema for both tools. Specify that responses contain a 'content' array with 'type: text' and 'text' field containing formatted news items. Include sample field names (title, link, pubDate, creator, content).
• Add min/max constraints to search-news 'days' parameter. Update schema to z.number().min(1).max(365).default(7) and add description: 'Number of days to search back (1-365, default 7).'
• Improve error messages with recovery guidance. Instead of 'Error fetching daily news', return 'Failed to fetch The Verge RSS feed. The service may be temporarily unavailable. Try again in a few moments.' Include retry hints for transient failures.

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

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