Open Comet
The autonomous AI browser agent for deep research & tasks
149 followers
The autonomous AI browser agent for deep research & tasks
149 followers
Open Comet is a high-fidelity AI agent that lives in your browser’s sidepanel. Unlike basic assistants, it autonomously browses, researches, and executes multi-step workflows across any website. Built on a "Zero-Data" architecture, it keeps your history local while providing enterprise-grade reasoning. Features include high-fidelity execution guards, STORM-inspired research loops, and full support for both Cloud (BYOK) and Local (Ollama) models.
Tried a few browser agents before and the scariest moment is always when it hovers over a "Submit Order" button. Does this one ask before doing something irreversible or do you just trust it and hope?
Open Comet
@dklymentiev That’s a very real concern, Dmytro 😄 totally get that “hover over submit” moment.
Open Comet is designed to not blindly execute irreversible actions. It follows a human-in-the-loop approach, where:
It asks for confirmation before any high-impact or irreversible action (like submissions, payments, etc.)
You can see what it’s about to do before it happens
All actions are transparent and trackable, not hidden
So the idea is—you’re always in control, and the agent assists rather than takes risky decisions on its own.
Building that trust layer is a big priority, especially for real-world use cases 👍
@princechouhan06 That's great to hear! Happy to test it!
every autonomous browser agent I've tried hits the same wall eventually - persistent auth sessions and bot detection. curious where this one breaks.
Open Comet
That’s a very real observation, Mykola—completely agree 👍
Persistent auth sessions and bot detection are definitely the hard boundary for most browser agents, and Open Comet isn’t magically bypassing those.
Right now, it works within your existing browser session, so it leverages your active login state rather than trying to handle auth itself. That avoids a lot of friction, but:
It doesn’t try to bypass bot detection systems
Performance can vary on platforms with aggressive anti-automation measures
Some flows may break when sites heavily rely on dynamic tokens or strict interaction patterns
So the current approach is more about assisted interaction within real user context, not full autonomous control across all sites.
Long term, I’m exploring ways to make it more robust around session handling and reliability, but staying within safe and compliant boundaries is important here.
Would be really interested to hear where it breaks for you if you try it—that feedback is gold 🙌
working within your existing session is the smart call - fingerprint spoofing just kicks off an arms race you can’t win
Git repo git clone https://github.com/princechouhan19/open-comet not found
Open Comet
@mike_evdokimov Ahh yes — let me clarify properly.
The public open-source repo (v1.1.0) is here 👇
👉 https://github.com/princechouhan19/OpenCometAI
The open-comet repo is currently private (v3.0.0 / unstable), so it won’t be accessible right now.
Thanks for pointing that out 🙌
really appreciate the zero-data architecture and local history storage. most ai tools are just black boxes with your data so keeping it in the browser is a big plus for privacy. definitely going to give the chrome extension a spin today
Open Comet
@yp_pro Really appreciate that, Yaroslav 🙌
Privacy was a core focus while building Open Comet — keeping things local-first and transparent instead of a black-box approach.
Glad that resonated with you. Would love to hear your thoughts after you try the extension 👍