Your July 13 Show HN got 1 point and 0 comments. Meanwhile, people have been asking for exactly this — an agent driving a real, logged-in, local browser — on Reddit, HN and the incumbent's own issue tracker. Every quote links to its source. Your launch wasn't the verdict — the room was just empty. These rooms aren't.
“What's the best setup for letting Claude Code use my logged-in Chrome? I've seen Browser Use / Browser Harness, Playwright MCP, dev-browser, and agent-browser mentioned. What actually works out of the box?”
A literal, verbatim ask for Narada's exact job — actively comparing tools, with 14 comments of people sharing the same need.
“Web browsers were never designed for agents… an agent needs a different environment: isolated workspaces, persistent logged-in state, parallel execution, and the ability to script complex web flows with code.”
Independently articulates your thesis word-for-word after failing real tasks with Claude Code/Codex agents.
“It feels like building a serious, long-running agent requires duct-taping a highly fragmented stack… Browserbase for the headless infrastructure, proxies, and session management… have you found a more unified runtime that handles this under one control plane?”
Architecting a production agent right now; names the cloud incumbent as duct-tape and asks for the unified alternative.
“For running a few browsers and GUI apps, you're much better off buying a Mac Mini than paying for a more expensive and worse-performing container in the cloud. Browsers were not designed to run in Linux containers.”
Already spending money to run agent browsers locally instead of the cloud — your wedge, three days before your launch.
“Having a hopefully seamless way to perform some automations in the browser on my behalf but since it's the browser I'm logged in to, it just makes the handoff between myself and the agent feel more seamless.”
Wants agent automations in his own daily logged-in browser — not a test harness. Shopping in public on a rival launch thread.
“The useful version of this for me is less 'agent can browse a website' and more 'agent can keep the browser state in the same work loop as the repo.' A lot of real tasks are half local files, half logged-in web UI, and the handoff between those two is usually where the work gets weird.”
Describes the local-context + logged-in-web gap on OpenAI's own launch thread — the gap cloud automation structurally can't close.
“I had an issue with Playwright MCP where only one Claude Code instance could be using it at a time… I also find it more useful that the Chrome MCP uses my current profile since I might want Claude to look at some page I'm already logged in to.”
Churned off the incumbent over concurrency; wants current-profile logged-in browsing. Three of your selling points in one comment.
“The cloud browser problem. My VPS is a small instance. Launching a real browser there — even headless — eats RAM I don't have… The login problem… fighting OAuth flows through a terminal, no saved sessions. It's miserable.”
Hit your exact two pains so hard he built his own bridge. Build-vs-buy pressure this strong means budget exists.
“Currently, each MCP connection gets one browser context that is ephemeral. AI agents that need to test authenticated flows must re-login on every session start… Re-login to multiple services on every test run.”
Persistent logged-in agent sessions, requested directly from the biggest incumbent — and closed without delivery. Five comments of agreement.
“We want them [Claude managed agent sessions] to have Playwright access to actually check their work in a browser… using --isolated to keep separate sessions loosely decoupled… has had a variety of issues like the browser state reverting on subsequent calls.”
A team running FLEETS of Claude agents hitting per-agent browser-session walls 17 days ago. That's a company-sized need, not a hobbyist one.
Imagine it every Monday morning: where your buyers asked for your product last week, with reply-ready angles. Free while we onboard the first founders — just send your product URL.
Get a radar for your product →