Two tools present with significant quality gaps. search_bid_announcements has a Korean description that is non-actionable for English-speaking LLMs, lacks parameter descriptions despite accepting two date parameters, and provides no output schema documentation. The hello tool has a reasonable description but its schema is overly complex (nested object with title/description fields) without clear use case or output documentation. Neither tool documents what they return, making it impossible for agents to chain results. No error handling guidance. Parameter descriptions are missing entirely for search_bid_announcements, violating the critical requirement that every parameter must have a description.

• Replace Korean description with English: 'Search for government procurement bid announcements by date range. Returns list of bid items with details like bid number, status, institution, and deadline.'
• Add parameter descriptions: bidStartDate: 'Bid announcement start date in YYYYMMDDHHMM format (e.g., 202401151430 for Jan 15, 2024 at 2:30 PM). Inclusive.' bidEndDate: 'Bid announcement end date in YYYYMMDDHHMM format. Inclusive.'
• Document output schema for search_bid_announcements: 'Returns array of bid items, each containing: bidNtceNo (string), bidNtceNm (string), bidNtceSttusNm (string), bidClseDate (string), asignBdgtAmt (string), ntceInsttNm (string).'
• Add error handling guidance: 'If the API returns an error, the service may be temporarily unavailable. Retry after 30 seconds. If date format is invalid, ensure YYYYMMDDHHMM format with valid month (01-12) and hour (00-23).'
• Clarify hello tool purpose: either remove it (appears to be a test tool) or document its actual use case, expected output, and when an LLM should invoke it vs other communication tools.