Company Scan Assistant

Problem

Screening companies with conditions like “Shanghai, post-B round, biomedicine” usually requires repeated keyword edits, region expansion, and financing amount normalization, while results are scattered across multiple sources. Company Scan Assistant turns that natural-language request into one structured search, reducing manual query assembly.

How It Works

The skill first parses intent and extracts parameters for industry, financing round, amount, time, and region. Domain terms are decomposed into no more than 10 concepts keywords covering both technology and product/application signals. Region phrases map to --provinces or --cities; when no region is given, mainland China is the default; amounts are normalized to ten-thousand yuan. It then calls search_companies.py with --output-temp, writing results to a system temp directory to avoid shell output truncation. After the script prints a confirmation path, the skill reads the file and presents Markdown results, including query summary, company counts, field table, and source links.

Boundaries

It depends on the Yiou database and network availability, and is best for batch company scans, competitor screening, or early deal sourcing. If results are empty, reduce keywords or relax round and region filters; overly broad keywords reduce precision, and the skill does not replace due diligence, registry verification, or detailed financial analysis.