AI WeChat Official Account Feed

Problem

AI official account posts update quickly and the volume is high. Reading hundreds of articles manually is impractical, and keyword search alone cannot reveal which topics are clustering on a given day. The need is a repeatable workflow for finding hot posts, grouping topics, and keeping a browsable archive.

How It Works

The skill scans AI WeChat official account content, with a default scope around 200 articles, then filters high-engagement posts using view counts, likes, comments, and other signals. It clusters the day’s content based on titles, tags, and topic terms, producing categories such as Agent, LLMs, and AI art. Output includes a terminal table and a dark-theme HTML report with cover images, authors, engagement data, article links, and date navigation. The report has built-in full-library search through a local proxy for fuzzy, date-independent lookups. Scope can be adjusted with --keywords, --count, and --date, and subscription can keep reports in a local folder.

Boundaries

It relies on a RedFox API Key, so usage depends on the registered account. Categories are determined by the day’s content and may change daily. Search is best for locating topic-specific articles across dates, while the report is a curated view of one day. If the default topics are narrow, add more keywords or increase the scan count.