National Bidding Data Automated Analysis Assistant

The Problem to Solve

Tender and bidding information is fragmented across regional and industry-specific platforms, with data formats varying widely and daily updates exceeding 300,000 entries. Manual analysis faces three core challenges:

• Information Overload & Missed Opportunities: Inefficiently covering all relevant announcements across the nation, specific sectors, or timeframes.
• Unstructured Data: Extracting key fields like procurement lists, models, and prices from narrative-heavy original notices for cross-project comparison is extremely time-consuming.
• Difficulty in Dynamic Monitoring: Setting up stable, automated processes to continuously track specific keywords, regions, or enterprises and perform trend analysis.

How This Skill Works

This skill encapsulates the complex logic of interacting with the Shibo Technology Bidding Data API, converting ambiguous analysis needs into precise data queries through natural language interaction. Its core workflow is:

1. Intent Recognition & Query Generation: When you pose a question (e.g., "search for smart city projects in Beijing from the last 30 days"), the skill parses the core requirements (keywords, region, time) and automatically supplements default conditions (e.g., announcement type "tender notice", excluding irrelevant industries).
2. Intelligent API Routing: Based on the need type (search, details, contract, planned projects), the skill automatically selects and invokes the most suitable API endpoint, such as searchProjectApi or SearchProjectForAI. It handles API authentication (BBIAO_API_KEY), pagination, and parameter assembly.
3. Structured Output & Integration: It consolidates raw API responses into clear tables or reports. Fields include Project Name, Region, Purchaser, Winner, Amount, and Link. For detail queries, it parses and highlights information like Procurement Requirements, Key Deadlines, and Attachments.
4. Error & Quota Handling: When the API returns insufficient balance (error code 0100590006) or authentication failure, the skill provides explicit guidance (e.g., a link to a recharge page) instead of blind retries. For empty results, it suggests expanding synonyms or relaxing query criteria.

Scope and Caveats

• Data Source Dependency: The skill's capabilities are entirely dependent on the API and data provided by Shibo Technology (99.99% coverage). Analysis results align with the quality and timeliness of the underlying data source.
• API Quotas and Costs: All calls consume API quota. For large-scale batch analysis or automated tasks, you must assess your package limits in advance.
• Automation Design: When building automated workflows, it is mandatory to include limits on query pagination, reasonable time windows, and explicit maximum retry counts to avoid infinite loops or unnecessary overhead.
• Field Constraints: Output strictly adheres to the 78 structured fields returned by the API. If the original announcement did not disclose certain information (e.g., specific contact details), the result will note "Not Disclosed" or "Unknown" and will never fabricate it.