Signal Radar

Find the story before the algorithm does.

Signal Radar reads Hacker News, Reddit, GitHub and a handful of feeds, and ranks every story against the topics you follow. There is no engagement feed and nothing is recommended to you. You see what is moving, and you decide what is worth making something about.

Reading needs no account. Pro is ₹99 for 30 days, and new accounts get 7 days free.

On the radar right now

Loading the feed

Ranked for the default topics. Your radar follows the topics you pick.

From a feed to a draft

Four steps are free. The last two are Pro.

  1. Read the radar

    Choose the topics you follow, or write your own keywords. Every story from every source is scored on keyword match and momentum: points, stars and comments per hour. The ranking runs in your browser, so reading needs no account.

    The Read view: a search field, topic tabs, and a ranked list of stories with source, age, points and score.
  2. Search anything

    Search the feed for a word and the list narrows to what matches. If the feed has not seen it yet, Search everywhere asks the live sources directly.

    The search field holding the word claude, with matching stories listed below and a link to search everywhere.
  3. Summarize without opening

    Summarize reads the page and gives you a short summary and a few points, so you can tell whether it is worth the tab. It is machine-written and says so under every summary. A free account gets 3 new summaries a day; a page someone already summarized costs nothing.

    A story with its summary open: a two-line summary and three bullet points, then a note that it can miss things.
  4. Save it to Ideas

    Save any story from Read, or paste a link, and it goes to your Ideas queue. Each run also adds new stories ranked for your topics. Pick what is worth making something about, skip the rest, and undo either.

    The Ideas page with a field to add a link, and a queue of stories each with Pick and Skip buttons.
  5. A hook, an angle, a title (Pro)

    On Pro, each queued story arrives with a hook, an angle and a title draft, and the queue learns from what you pick and skip. They are drafts. You edit them and post them yourself.

    A queued story with a written hook, an angle and a title draft under its summary.
  6. The digest comes to you (Pro)

    After each run, one message on Telegram and, if you switch it on, a push on your phone. If you keep ideas in a Notion database, Pro syncs the queue there too.

    The Notifications settings: web push switched on, and Telegram connected.

Free and Pro

Free

₹0

  • The ranked Read view across every source, no account needed
  • With an account: an Ideas queue of up to 10 new stories a run, ranked by your topics
  • Save any link to Ideas, up to 10 a day
  • 3 new page summaries a day; ones already read are free
Create a free account

Pro

₹99 for 30 days

Also ₹249 for 90 days, ₹799 for a year.

  • Everything in Free, with higher limits: 30 saved links and 100 summaries a day
  • A hook, an angle and a title draft written for each queued story
  • An Ideas queue that learns from what you pick and skip
  • Your own RSS feeds
  • A Telegram digest, web push and Notion sync
Choose a pass

Pay once. Time adds up: buying again adds days to the end.

Questions

What is free?

Reading. The ranked Read view, with search and every source, needs no account. A free account adds an Ideas queue of up to 10 new stories a run, 10 saved links a day and 3 new page summaries a day.

What does Pro cost?

₹99 for 30 days, ₹249 for 90 days, ₹799 for a year. No subscription and no auto-renew. Buying again adds days to the end. New accounts start with 7 days of Pro free.

Where do the stories come from?

Hacker News, Lobsters, DEV, Mastodon trending, Reddit and GitHub, plus RSS feeds: GamingOnLinux, The Verge, Product Hunt, Fireship on YouTube, Google News and Neovim releases. Pro lets you add your own feeds.

Does it post anywhere?

No. It never posts, replies or messages anyone but you. Hooks and titles are drafts for you to edit and publish yourself.

What does the AI do?

Ranking in Read is keyword match and momentum, with no AI. Page summaries and, on Pro, the scoring of your queue and the hooks, angles and title drafts come from a language model. It works from the story text, can be wrong, and does not choose your topics for you.

What happens to my data?

Topics and sources stay in your browser until you sign up. An account stores your email, topics, queue and saved links, and story text goes to the AI provider for summaries and scoring.

Can I run it myself?

Yes. It is open source under AGPL-3.0 and runs on Node and Postgres with your own keys. The code and setup steps are on GitHub.

See what is moving today.