About this skill

Problem

Tech media outlets publish many headlines, flash updates, and roundup posts every day. Manually filtering them makes it hard to separate actionable AI signals from background noise, false positives, and cross-source duplicates. Multi-day events also lack a clear trend line.

How It Works

AI-News-Collector uses a rule-based engine to collect articles from RSS sources such as 36Kr, QbitAI, InfoQ, and TMTPost without API keys. The core workflow includes:

Keyword and domain filtering: scores items around AI, funding, technical breakthroughs, market impact, and strategic value.
Contamination detection: identifies roundup posts and extracts only the sub-news item that matches the title.
Three-stage fallback: tries full-text fetch, search supplementation, then RSS summary to reduce link breakage.
Negation context: lowers false positives from phrases such as “no acquisition plan” or “denied rumors.”
Daily and weekly reports: produces a structured daily digest and maintains a 7-day view of topic heat, investment signals, active companies, and high-value themes.

Fit and Caveats

It fits a local Python setup for Chinese tech news automation, especially for AI, robotics, semiconductors, and funding tracking. If media sites change markup, RSS feeds fail, or the focus area shifts, adjust config.yaml and source settings. The rule engine is less suited to English foreign media, video content, or deep research reports.

Screenshots

Use Cases

Collect daily tech stories from 36Kr, QbitAI, InfoQ, and TMTPost into an archivable AI digest.
Run a local Python collector that filters non-AI noise and duplicate reports using keyword rules.
Track companies and events across days with a 7-day topic heat, investment signal, and company report.
Route generated Markdown daily digests into Obsidian for review by date, theme, and event chain.

Best For

AI researchers who read Chinese tech media daily and want automatic noise filtering plus event capture.
Data engineers maintaining local automation, focused on RSS fetch, summary cleanup, and report generation.
Investment analysts tracking AI, robotics, semiconductors, and funding signals in daily digests.
Knowledge managers who archive tech updates in Obsidian by daily, weekly, and event-chain views.