Policy Intelligence Radar

The problem

Government policy pages are updated frequently and are spread across multiple ministry and provincial agency sites. Manually checking national and Sichuan-related sites, filtering by terms such as AI, smart manufacturing, and industrial internet, then assembling a shareable daily brief is repetitive and easy to botch. finding turns that workflow into discrete, scriptable steps for teams that track policy changes on a daily basis.

How it works

The skill centers on three operations: collect_policy fetches policy items from configured national and local sources; generate_report filters by keywords and produces a plain-text daily report with titles, links, publication dates, issuing agencies, and summary statistics; send_to_wechat pushes the content to a WeCom group through a bot Webhook. A scheduled run defaults to 09:00 every morning, and manual triggering is also supported. Reports are written to ./reports/policy_report_YYYYMMDD.txt, which makes archiving and downstream reuse straightforward.

Limits to keep in mind

This skill fits use cases such as corporate strategy, investment research, project application, and industry trend analysis where continuous policy tracking matters. It is not a real-time news engine: results depend on target site stability, network reachability, and page update timing, so parsing rules should be reviewed periodically. WeCom messages are limited to 2000 characters, and overly long content is truncated; personal WeChat does not support Webhook push, so use a WeCom bot.