Recent activity
- PostedModel benchmarks become browser-state lotteries without a fixed execution layer
- PostedSearch is not evidence: a routing model for CLI research agents
- PostedHow do you verify a Copilot code change against the live UI?
- PostedWe ranked Claude Skills for AI agents, QA, scraping, research, and growth
- PostedBrowserAct vs Playwright: the real question is which browser layer you want to own
- PostedI got tired of rewriting the same scraper every month — so we open-sourced the tool that writes it once
- PostedMy $5,000/month X API bill, and the architecture I should have built first
- PostedMost "Claude skills" are prompt templates with delusions of grandeur — and the bill proves it
- PostedMy AI agent burned 180K tokens on a 30-line task. Here's where they all went.
- PostedBrowserAct hit Product Hunt #1. The bigger signal: agent execution is becoming infrastructure.
- PostedBrowserAct just became #1 Product of the Day on Product Hunt — now we’re pushing for the weekly leaderboard 🚀
- PostedMy Lead Said Browser Skills Were Just Prompts in Disguise. I Tested BrowserAct Anyway.
- PostedHow we stopped wrong-account mistakes in multi-account social automation
- Commented onTwo browser-agent failures we kept seeing: login and infrastructure mismatch
- Commented onTwo browser-agent failures we kept seeing: login and infrastructure mismatch
- Commented onTwo browser-agent failures we kept seeing: login and infrastructure mismatch
- Commented onTwo browser-agent failures we kept seeing: login and infrastructure mismatch
- PostedTwo browser-agent failures we kept seeing: login and infrastructure mismatch
- Commented onA Google Maps scraper is not enough. The pipeline matters more.
- PostedA Google Maps scraper is not enough. The pipeline matters more.