OPEN_SOURCE ↗
GH · GITHUB// 19d agoOPENSOURCE RELEASE
Last30Days Skill turns chatter into briefs
mvanhorn's last30days is a Claude Code skill that pulls recent signals from Reddit, X, YouTube, Hacker News, Polymarket, and the web, then synthesizes them into cited research briefs. The current v2.9.5 README shows it has already expanded beyond the scraped blurb into Bluesky, Truth Social, TikTok, Instagram, comparative mode, and per-project config.
// ANALYSIS
Hot take: this is less a search skill than a recency layer for AI research, and that is the interesting part. The moat is not any single source, but the cross-platform synthesis of what communities are actually saying right now.
- –v2.9.5 broadens the source net to Bluesky, Truth Social, TikTok, and Instagram, which makes it feel more like an intel engine than a Reddit/X sampler.
- –Comparative mode is the smartest product move here because "A vs B" questions usually need synthesis and judgment, not raw retrieval.
- –Auto-saving briefings to `~/Documents/Last30Days/` turns one-off queries into a reusable knowledge base, which is a useful workflow moat.
- –The tradeoff is latency and operational overhead: this will beat chat search when depth and provenance matter, but not when someone wants a fast answer.
- –For AI developers, the best use case is spotting prompt patterns, tool recommendations, and emerging workflows before they ossify into stale blog-post lore.
// TAGS
last30daysagentsearchresearchcliopen-sourceautomation
DISCOVERED
19d ago
2026-03-24
PUBLISHED
19d ago
2026-03-24
RELEVANCE
9/ 10