0.1.16 is about reading more of the web without writing anything: point Manycrawl at a page that lists things, check on the page itself what it picked up, and go behind a login when the list lives there.
List scraper and Detail scraper
Paste the address of any page that lists things, like a directory, a marketplace or a search result, and the List scraper turns it into a table with one row per item. It finds the list and its columns on its own, follows "next page" and "load more", and uses no AI, so a run costs nothing extra and gives the same result every time.
The Detail scraper goes one step further: it opens each row's own page (a profile, a listing) and adds what it finds there as new columns in the same table.
See what gets scraped, on the page
Show on page opens the page in Manycrawl's browser with every column outlined on it, each in its own colour and numbered to match the table. If a column picked up the wrong thing, you see it immediately. Click an outline to leave that column out or put it back.
Scrape behind a login
Some lists are only visible once you are signed in. Tick behind a login, sign in once in Manycrawl's browser, and the List and Detail scrapers read with that login from then on. Manycrawl never asks for or stores your password; the login stays on your computer, like it would in your own browser. If it expires, the read stops and asks you to log in again rather than filling your table with empty rows.
Classify
Classify is a new column for questions with a fixed kind of answer: yes or no, pick one of these options, or a score. Every answer comes with a confidence value, so you can filter down to the rows the model was sure about. It needs a TypeSafe API key.
AI columns, up to 16× faster
AI columns now work on several rows at once. On a 298-row table, a run that took 400 seconds now takes 25. Classify has a "Parallel rows" setting if your API key has tight rate limits.
Cleaner Google Maps results
In Google Maps tables, a business's Website is now always its real site. Reservation links from booking services go into their own Booking link column instead, and email and scrape columns read the real Website by default, so they find the business's emails rather than a booking widget's.
Maps searches are also more dependable. When Google stops cooperating with a search window, Manycrawl now resets it and tries again, and a search that genuinely finds nothing says so instead of just returning an empty table.
For AI agents
If you drive Manycrawl from Claude, Codex or another agent over MCP, two new tools:
- Stop: an agent can halt a running column or search, the same way the Stop button does.
- Read the log: an agent can read the Activity Log for a table, so it can see why a run failed or skipped rows, not just that it did.
Table status now shows the run in progress and, when a search fails, the reason.
Get it
Already installed? Manycrawl updates itself, so 0.1.16 arrives the next time the app starts. New here: download for Mac or Windows.
Try the List scraper on a page you already use
Free plan includes 500 actions a month, unlimited tables and rows. No card.