Telegram Digest

To aggregate Telegram posts into a digest, this Skill extracts public posts from configured channels, scores them by views and reactions, and clusters related items into narratives. It fetches recent content via WebFetch and outputs a bounded-length notification.

The best way to filter noise and deduplicate Telegram content is using signal scoring and memory logs. This Skill removes low-signal text, promos, and bots, while checking URL-level memory logs to prevent re-surfacing items already seen in the last 2 days.

Yes, you can automate cross-channel Telegram monitoring for morning briefs. This Skill operates autonomously to fetch posts, compose a narrative-based digest, send it via notifications, and append a run log for traceability without manual intervention.

Narrative clustering for Telegram aggregation works by grouping related posts across different channels into unified stories. Instead of generating a channel-by-channel feed, it ranks and clusters high-signal content based on views, reactions, replies, and link presence.

Yes, you need to configure specific channel inputs to generate a Telegram news digest. You must provide target sources via a var or channels.md file so the Skill can perform best-effort fetching, extract posts, and apply filtering for your digest.