Product · Browser on the board

The web, on the board.

Your agent opens real pages right on the board, reads them, highlights what matters and pins notes to them. Your team browses the same pages alongside it.

What your agent does with a page

The same moves you'd make with twenty tabs open, except everyone can see them and nothing gets lost when the tabs close.

  • Open or search

    Opens a link you give it, or searches and opens the results worth reading.
  • Read

    Reads the real page, not a summary someone pasted into a chat.
  • Click and scroll

    Follows links, scrolls, goes back and forward, and refreshes.
  • Find on the page

    Jumps straight to the line that answers the question.
  • Highlight

    Marks the section that matters, in place, for everyone.
  • Pin notes

    Sticks a note to a section so the why sits next to the what.
  • Collect images

    Lists a page's images so the useful ones end up on the board.
  • Snapshot

    Freezes the page as the agent saw it, even if the site changes later.

Highlights everyone sees

When the agent marks a line, the whole team sees the same mark on the same page. Add your own next to it. Highlights and notes stay put when anyone reopens the board.
  • Agent and teammate highlights side by side, each in its own color
  • Notes pinned to the exact section they're about
  • Restored when the board reloads, so the research doesn't evaporate

You browse too

Browser cards aren't a recording of the agent's session. They're live pages anyone on the team can scroll, click through and mark up while the agent works in the next card over.
  • Several pages open on one board, side by side
  • Collect a page's images for the board
  • Your notes and the agent's share one board

Proof of what it read

Sites change. A snapshot freezes the page exactly as the agent saw it, so a plan built on last week's numbers still shows last week's numbers.
  • Freeze any browser card as a snapshot
  • Check the agent's claims against the page it actually read

Collect the good bits

Venues, products, competitors' screenshots: the agent can list a page's images so the right ones land on the board, next to the note that explains why they matter.

What it does, and what it won't

The agent researches. People act. That line is built into the product.

Your agent can

  • Open links and search the web
  • Read pages, click links, scroll, go back and forward
  • Highlight sections and pin notes to them
  • List a page's images
  • Freeze a page as a snapshot

It never

  • Signs in to your accounts
  • Types into forms
  • Buys, pays for or reserves anything
  • Acts on a site on your behalf

Questions about the browser

Short answers about what loads where and who can do what.

Can the agent sign in to sites or buy things for me?

No. It opens, reads, clicks, scrolls, highlights and annotates pages. It never logs in, types into forms, pays or reserves anything. When it's time to act, a person does it.

Where do the pages load?

On your machine, not on our servers. In the desktop app, browser cards are a full built-in browser. In a web browser, pages load through the local host you run with npx ideawave. Highlights, notes and snapshots are saved with the board so your team sees them.

Can my teammates use the browser cards too?

Yes. Editors can browse in the same cards, add their own highlights and pin their own notes. Viewers see everything live but can't change it.

Do I need the desktop app?

Browser cards work best there, and it needs no terminal. The desktop app runs on macOS (Apple Silicon) and Linux. On Windows, or if you'd rather not install it, run npx ideawave instead.

Put the web on the board.

Browser cards work best in the desktop app for macOS and Linux.