HKEx Filing Bulk Search And Download

Problem

In HK stock audit, research, and compliance work, teams often need to verify HKEx filings by stock code and date range. Manually opening the web search page, copying links, and downloading PDF/HTML files is slow and can miss paginated records; preparing work papers then requires adding company names, dates, file types, and hyperlinks.

How It Works

The skill calls the JSON API behind the HKEx title search page: https://www1.hkexnews.hk/search/titleSearchServlet.do. It does not require a browser and uses plain HTTP requests to query filing metadata with --stock, --from, --to, and --limit, with output available as json or xlsx.

Key steps include:
- Search filings: query by 5-digit stock code and date range, defaulting to the last two months.
- Bulk download: the download subcommand fetches announcement PDF/HTML files in parallel, with --workers 5 by default.
- Metadata export: fields include stockCode, stockName, title, date, link, fileType, fileSize, and filingType.
- Filing classification: records are grouped as ANNUAL_REPORT, RESULTS, CIRCULAR, NOTICE, or OTHER, making annual reports, interim results, and circulars easier to filter.

Boundaries

The tool fits audit work papers, disclosure completeness checks, and filing trend comparison, but it depends on an unpublished HKEx endpoint. The API path, fields, or rate limits may change. All-market or long date-range queries can return many records, so start with a small --limit sample; local downloads should still respect the source website’s terms of use.