jobscope-mcp has well-structured tool definitions with clear naming (verb_noun pattern) and comprehensive input schemas using Zod. All four tools have descriptions and typed parameters with enums where appropriate (e.g., remote filter). However, parameter descriptions are sparse or missing in several cases, output schemas are not documented, and error handling lacks recovery guidance. The server follows basic MCP patterns but falls short of production-grade polish expected for A-tier tools.

• Document output schemas for all tools. Define the structure of search_jobs results (array of {id, title, company, location, url, ...}), get_job details, list_companies response, and list_sources response. Include field types and descriptions.
• Expand parameter descriptions. For 'ats' array, explain valid ATS platform names (e.g., 'Greenhouse, Lever, Ashby'). For 'posted_since', document format: '30d' = last 30 days, '7d' = last 7 days, or null for all. For 'companies', clarify that slugs come from list_companies.
• Add recovery guidance to error responses. Categorize errors as retryable (network timeout, rate limit) vs user-fixable (invalid company slug, malformed query). Return actionable messages: 'Company slug not found. Call list_companies to see available companies.'
• Enhance list_sources description. Explain that this tool reveals available ATS platforms before calling search_jobs. Current description (60 chars) is too brief. Target 100-150 chars with context.
• Document pagination behavior. Add to search_jobs and list_companies descriptions: 'Results are limited to the specified limit (default 50, max 200). Call again with offset or cursor to fetch additional results.' Return a next_cursor or total_count in responses.