SMF·BARDEEN
Can a prompt replace Bardeen?
Automation — workflow automation and app integrations
Exhibit tracking slip
Verdict
A browser extension that can scrape structured data off the page you're viewing and send it to another service — Bardeen's actual trick, working even on sites without an API — is a real weekend build using a content script and a fixed set of two or three site/service pairs. What doesn't survive: Bardeen's library of hundreds of pre-built 'playbooks' across many site and service combinations, and reliability as each site's page structure quietly changes.
Exhibit A — The prompt
Received on31.07.2026Build a browser extension that extracts data from the page you're currently viewing and sends it to one connected service — working on sites that don't have their own API, which is Bardeen's actual differentiator over a pure API-to-API tool like this catalogue's Pipedream entry. Use a Chrome extension (Manifest V3) with a content script that, when triggered by a keyboard shortcut or a right-click context-menu item, extracts structured data from the current page using CSS selectors you define per site, such as product name, price, and URL from an e-commerce page. Send the extracted data to one connected service's API — a spreadsheet, a note-taking app, or a CRM, pick one — using that service's documented API and OAuth or API key. Show a confirmation preview of the extracted data before sending, so a broken selector from a site redesign is caught before bad data gets sent, not after. Store a small library of site-specific extraction configs, which CSS selectors map to which fields per domain, locally, editable when a site changes and breaks extraction. Do not build a library of many pre-built playbooks across dozens of sites and services, or a visual no-code builder — those are out of scope; this handles exactly the site/service combinations you configure yourself. Needs a domain for OAuth callback handling and an OAuth app registration with your chosen destination service.
Opening prefills the prompt — press enter to run it.
Exhibit B — What you lose
- B.1 hundreds of pre-built site/service automation playbooks
- B.2 reliability as target sites' page structure changes over time
- B.3 a visual no-code automation builder
- B.4 team-shared automation libraries
Prior art
Exhibit C — Why people still pay: connector breadth and execution reliability
Scraping one page's structure once is easy; maintaining working extraction logic across hundreds of sites that constantly redesign their pages, without you noticing something silently broke, is the maintenance burden.
Questions
Does it work on sites without their own API?
Yes — that's the core mechanic. It reads structured data directly from the page's HTML using CSS selectors, which works regardless of whether the site itself offers an API.
What happens when a site redesigns its page and breaks the selectors?
Extraction will silently produce wrong or empty data until you notice and update the selector config — there's a confirmation preview specifically to catch this before bad data gets sent, but ongoing maintenance is still on you.
How many site/service combinations does it support out of the box?
None pre-built — you configure each one yourself. Bardeen's library of hundreds of ready-made playbooks is real, accumulated product work this build doesn't include.
What does it cost to run?
Minimal hosting for OAuth callback handling plus whatever the destination service itself costs — no per-automation fee.
Related tools
Receipt