Zhiliao Tender Data Assistant

Problem to Solve

Tender data is often scattered across announcement text, company registry fields, winning entities, brands, and price records. Engineers searching bids need to combine keywords, regions, budgets, stages, purchaser or winner constraints, and then connect competitor and price history. Raw data assembly can miss subsidiaries, confuse brands with purchased items, or blur the roles of purchaser, bidder, and winner.

Core Capabilities and Workflow

• Bid search: use search_bids and query_bids_advanced for structured filtering, then call get_bid_detail for full text.
• Company analysis: use search_company to resolve entities, get_company_profile for registry and bidding activity, and find_competitors to locate overlapping competitors.
• Market and price: use get_top_purchasers, get_top_suppliers, get_top_brands, and get_price_trends to inspect buyers, winners, brands, and historical unit prices.
  Key parameter is match_modes: use caller for projects published by a company, winner / tender for winning or bidding projects, brand for brand search, and fulltext as fallback. Complex conditions can combine keywords, keyword_groups, and exclude_keywords, while bid_process filters purchase intention, tender, award, contract, and related stages.

Boundaries and Notes

This skill is useful for objective bid lookup, entity mapping, top-brand analysis, and price trend organization. It does not replace financial audit, legal due diligence, or unverified market forecasts. Calls are subject to quota and rate limits: check the key on AUTHENTICATION_FAILED, slow down on RATE_LIMITED, and remember that keyword_groups requires the advanced search endpoint. For policy, strategy, or industry structure, prefer tender data first and supplement with reliable external sources.