Architecture Bid Insight: Zhulong Bid Insights

Problem

Construction tender data is scattered across announcements, projects, company records, and price history. A simple keyword search rarely answers practical questions: what a company procured, who won the bid, which projects are nearing expiration, or how a brand and model has been priced over time. This skill turns the zhiliaobiaoxun tender API into an executable query workflow, reducing the need to stitch together raw announcements, company data, and market statistics.

How It Works

It covers 16 tools across three groups:
- Bid search: search_bids and query_bids_advanced support keywords, region, amount, time, and advanced logic with keyword_groups and exclude_keywords
- Company analysis: search_company, get_company_profile, get_company_contacts, and find_competitors help resolve company entities, view business profiles, locate contacts, and identify competitors
- Market analysis: get_top_purchasers, get_top_suppliers, get_top_brands, and get_price_trends track purchasing units, winning suppliers, brand models, and price history

A key concept is match_modes: caller searches the purchasing party, winner searches the winning supplier, and tender searches bidders. For company-specific queries, keyword matching alone is not enough; the matching field must be explicit. The API defaults to stages such as procurement intent, pre-tender, tender, award, and contract, which can be narrowed using bid_process.

Notes

The skill depends on an API key or trial account, and quota exhaustion is handled by source-specific prompts. keyword_groups requires query_bids_advanced; the standard search endpoint does not support all advanced logic. Web information is useful for company background, trend context, and industry judgment, but the tender API responses should remain the primary data source.