Hacker Newsnew | past | comments | ask | show | jobs | submit | thawab's commentslogin

Thanks a lot, i wanted to ask about the headless agent use case: How does it compare to using https://github.com/vercel-labs/agent-browser

BrowserOS should work with agent browser as well in headless mode!

On top of that, if you want headful mode, you can use our MCP server https://docs.browseros.com/features/use-with-claude-code

Would love to understand your use case! You can hit me up at nithin[at]browseros.com


After getting burned by Watson. I am not touching any AI from IBM.


Tell us more about how you were burned, please!


It's a file you can run on your computer


How can o4-mini be at 57, and sonnet-4 is at 39? This is way off, o4-mini is not even in the top 5 of coding agents.


Does facebook use llamas in their model? it's a name, it doesn't have to be 100% true to its meaning.


There are several videos in the link that does this.


Where?! I only see intro videos to each lesson. No actual "lesson" being shown.


Start course. It’s an mcp, talk to it like an instructor.


3. Draw the rest of the fucking owl

This is not enough information for someone new to MCP / agents to start the course. What you actually need to point out, is that users need to type "startMastraCourse" into their IDE's agent mode to actually start the course.


I'm brand new to MCP and agents but was able to read the extra docs to get VSCode set up with Mastra. Then what? I only figured out the "start Mastra Course" because of their tweet where they show someone typing that into co-pilot.

There really needs to be more hand holding to get someone to the point where the course actually starts. From there I've been able to follow along alright, but it was a real battle to get to this point.


i hear you, we should have added more instructions since it's a new concept (and still very experimental)

we will update the top intro video to show how to actually get into the course and add more information to the course page to make this clear


are the Llama 4 issues fixed? what is it good at? coding is out of the window after the updated R1.


Yes, the issues were fixed ~1-2 weeks after release. It's a good "all-rounder" model, best compared to 4o. Good multilingual capabilities, even in languages not specifically highlighted. Fast to run inference on it. Code is not one of its strong suits at all.


Yes, a lot MCP servers are just api wrappers. Airweave looks like it copies the data and has a RAG that is processing your queries.


Hello Joe, thanks a lot for hydra and pg_duckdb. I wanted to confirm that for self hosting hydra i have to generate a token from your platform? what data is shared with hydra for this case. We need to double check as our data has restriction of sharing.

> Visit http://platform.hydra.so/token to fetch the access token and paste it into the section above.


Hello thawab, yes! you can self-host Hydra with a token from the platform. Sign-up and visit that URL to take you to the right spot. We call it Bare Metal deployment, here's 1 minute setup guide (https://docs.hydra.so/guides/bare_metal)


thanks a lot, the other part of the question:

1- what data is shared with hydra for this case?

2- whats the pricing for the bare metal deployment?


billing (usage) metrics so we know what to charge. We offer BYOC 'Bare Metal' deployments as part of the Business plan. You can set it up now, but we offer volume discounts so you should talk to our team directly. Feel free to DM me on X (@JoeSciarrino) or email founders@


for my daily work i use aider, but i jump into roo and client when there is an update, or someone has an interesting workflow. it's like taking a test drive.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: