Hash is an 🏷️#ai-powered personal news-📝synthesis engine that reads every source you follow — RSS feeds, email newsletters, forwarded emails, 📝Reddit, 📝Hacker News — and delivers one personalized daily digest and one podcast episode, anchored in what you already know. Positioned as "the original 📝MythOS · since 2017," its promise is simple: stop re-reading, start re-thinking.
How It Works
Hash runs a three-stage daily pipeline: ingest → synthesize → deliver.
- Ingest — Sources are scraped and polled around the clock. Multi-story newsletters are automatically split into individual stories (images and links preserved, attachments skipped), and a rolling 30-day story 📝manifest deduplicates coverage so old stories never resurface as new.
- Synthesize — Each morning Hash selects the most synthesis-worthy stories from the accumulated content, then writes a tiered email digest and a conversational podcast transcript. Perspective injection from the reader's 📝MythOS library anchors both outputs in their own prior writing — the news interpreted through your lens, not a generic one.
- Deliver — By 6 AM the digest lands in your inbox and the podcast episode publishes in a custom voice. Episodes have public pages, and public shows accept subscribers, so a synthesis built for one mind can be shared with 📝many.
Shows and the Feed
Content is organized into shows, each with its own sources, output modality (podcast or digest), and voice profile. Alongside the daily outputs, a curated feed surfaces the selected stories individually — and from any story, a reader can save it into their MythOS library as a reference memo or initiate a research brief that cites sources for every claim.
Closing the Loop
Hash is the intake half of a feedback loop with MythOS: what you find in Hash becomes 📝memory in MythOS, and what you've written in MythOS shapes 📝how Hash synthesizes tomorrow's news. 📝Memory → Mind → Mouth → Myth, running as a daily engine.
Plans
- Scribe (free) — 1 show, 20 sources, daily digest, 30-day history
- Scholar ($12/mo) — 3 shows, 100 sources each, custom voice, research capability
- 📝Oracle ($39/mo) — unlimited shows, team library, weekly deep research, priority synthesis