No more diving into complex DOM structures or writing fragile XPath expressions. Just specify what data you are scraping from the web with natural language-like queries, and AgentQL handles the rest.
Unblocked AI Code Review — High-signal comments based on your team's context
High-signal comments based on your team's context
Promoted
A huge steps towards creating Agent workflows! Now with accurate LLM assisted data extraction and webpage interaction, one can only image what web-agents are capable of to make daily life so much easier!
My fellow data detectives, number crunchers, and info wranglers - remember those days of endless clicking through websites, hunting for elusive info? I do. Whether chasing sources as a journalist, crunching numbers in finance, or piecing together market intelligence in consulting, data gathering was always a challenge.
That's why I am excited as both a user and a team member of AgentQL. We are building a tool that automates repetitive web interactions, e.g., point, click, scroll and get data from web or app UI. The nice thing is it uses an accessible query language that even I can learn :) It simplifies gathering and structuring web data for various projects.
👀 Please check out our technology at AgentQL.com
AgentQL isn't just another tool; it's a new paradigm for interacting with content on the web that addresses real, persistent challenges in the field. Full disclosure: I worked with the team to reach out to developers and learn what they needed to succeed. And it is because of the team's swift responses to the pain faced by developers that I believe in this tool. This team wants developers to win.
When I first tried out AgentQL, it felt like magic. I have written a lot of selectors in my life: CSS, jQuery, JavaScript. I know how to make specificity work, and I know the pain of that house of cards collapsing for want of a single id. AgentQL made it so easy to get an element and its contents using plain English. Since then, I can't stop finding uses for it in my workflows!
AgentQL has the potential to change how we build for the web. The secret sauce is AgentQL's AI-powered smart selectors. They're resiliant to UI changes, which we have the least control over and a major pain point.
It's refreshing to write queries in natural language and have them work across different sites with similar structures, saving countless hours of tweaking and maintaining scripts.
The Chrome extension is great for debugging and perfecting queries in real-time, but it's also my preferred way to grab all the social links off a speaker's website!
I also appreciate the flexibility in defining custom data structures. It gives me precise control over the output, which is crucial when integrating with other parts of my data pipeline.
The Playwright integration is the cherry on top for those of us working on end-to-end testing or automation. It's seamless, easy, and powerful.
Kudos to the Tiny Fish team for creating something that genuinely pushes the boundaries of what's possible for developers on the web. I'm excited to see how AgentQL evolves and the impact it will have on our industry.
We're all about pushing boundaries and making things easier for developers on the web. It's exciting to think about how AgentQL will evolve and the impact it could have on the industry. We’re thrilled to have the community on this journey with us!
Cool! I'm so needed! It seems that it can execute many automated processes on websites. I want to know if it can support X and LinkedIn. Will data scraping be prohibited by some websites?
That's a really great question, Zenda! We currently don't have first party solutions directly for bot or scraper detection / avoidance. However, since our SDK is currently an extension to playwright, there are a of compatible tools in and around the ecosystem that might be able to help! There are a few themes of techniques for this-- implementing an IP proxy rotation, using real / rotating user agents (supported by playwright!), and rate limiting interaction speed. You can also try doing this through our enable_stealth_mode API in the SDK (https://docs.agentql.com/api-ref...).
Other options to explore include using Chrome CDP or similar (example here: https://github.com/tinyfish-io/f...), re-using trusted session credentials, or some other tools that directly are compatible with playwright 🙂 If you find another solution or have other recommendations, we've love to hear about how you tackle this problem!
@zenda1122 Glad you found us! X and LinkedIn are both supported by AgentQL. In fact, I built an agent myself that automatically adds LinkedIn profiles of meeting attendees to my calendar—it’s been a huge time-saver!
Just had the chance to plug AgentQL into my personal web automation work flow and it worked wonders. Only if I had this 2 months ago I would've saved countless hours DOM searching and fiddling with playwright. Looking forward to what the future has in store for AgentQL and TinyFish!!!
Report
@zechereh Same here it has saved me so much time. Projects that took me weeks I can do in a couple days now
Hey Zachary! Great to hear you tried it out and liked it. Lets make it better together? You share your feedback and pain points, and we will address them and roll out even stronger!
AgentQL
AgentQL
AgentQL
AgentQL
AgentQL
AgentQL
AgentQL
LinkedCRM AI
AgentQL
AgentQL
AgentQL
AgentQL
AgentQL
AgentQL