Your attention,
your rules.
The web is full of algorithms designed to capture your attention and sell it. /feed gives it back — a personal, private reader that puts you in control of what you read, when you read it, and why.
The problem
You didn't choose
what you read today.
Every major platform runs on the same model: capture attention, maximise engagement, monetise the gap between what you want to read and what keeps you scrolling.
The algorithm is not your friend. It's a system optimised for someone else's revenue — not your clarity, not your time, not your thinking.
/feed is built on a different premise. You choose your sources. You set the filters. You own the feed. No recommendations, no engagement traps, no third party between you and the information you actually want.
How it works
Add your sources
Any website, blog, newsletter, or publication with an RSS or Atom feed. Add as many as you want — no arbitrary limits. Your list, your logic.
Organise and filter
Group feeds into categories, apply tags, save entries for later. Full-text search across everything you've ever subscribed to.
Read on your terms
Clean reading view, no tracking, no ads. Your feed is yours — export it, back it up, connect it to any LLM client via MCP.
What's inside
First-class AI tooling
Connect Claude, Cursor, Zed, or any MCP client to your personal feed endpoint. Your reading list as a data source for any LLM.
No tracking. Ever.
Optional Tor routing for feed fetches. We don't profile you. No ads, no engagement model, no attention sold.
Any source, any client
RSS, Atom, Fever API, Miniflux-compatible API, OPML import/export. Connect Reeder, NetNewsWire, ReadKit — or the CLI.
No more truncated summaries
Per-feed scraper fetches and extracts full article text. CSS-selector rules and Readability fallback included.
See it your way
Multi-column dashboards, filter rules, and smart views surface what matters and mute the noise — automatically.
Understand your habits
Heatmap calendar, reading streak, most-read feeds, best time to read. Actual useful data about how you consume information.
Bring your own AI
Plug your own LLM into your feed.
SlashFeed speaks the Model Context Protocol (MCP). Point Claude, Cursor, Zed — or any MCP client — at your own private endpoint and your reading list becomes a tool your AI can actually use: reading, searching, organising, and summarising on your behalf. Your AI, your data, your rules.
Not locked to one AI
Claude, Cursor, Zed, or anything that speaks MCP. Bring the assistant you already use — we don't ship a black-box AI you can't swap out.
Your endpoint, your data
Each user gets a personal endpoint scoped to their own feeds, secured with a personal token or OAuth. Nothing leaks across accounts.
Real tools, not just chat
List and search entries, subscribe to feeds, organise with decks and tags, mark read, save for later — the LLM drives the app for you.
Everywhere you read
One account, every surface.
Start in the browser, keep reading on your phone, subscribe with a click from your toolbar, or script it all from the terminal.
Web app
The full experience in any browser — installable as a PWA, works offline.
Available nowMobile
Native apps for iOS and Android with widgets, badges, and offline reading.
Coming soonExtensions & CLI
One-click subscribe for Chrome & Safari, plus command-line tools on GitHub.
Coming soonPricing
Free
$0
forever
- Up to 500 feeds
- Full reading experience
- MCP access
- OPML import/export
- Reading stats
Pro
$5
per month
- Unlimited feeds
- Full-content scraper
- Priority fetch
- Newsletter inboxes
- REST API + Fever API