Live data from Hacker News

Show HN: Paseo – Beautiful open-source coding agent interface

github.com

51–60 of 67 posts

Re: Show HN: Paseo – Beautiful open-source coding agent interface

#51

Nice work. I've been building a token usage dashboard for Claude Code — seeing how much each project actually costs. Would be curious if Paseo surfaces any token/cost visibility too.

I recommend https://github.com/wesm/agentsview - its quite good, has usage metrics and nice viewer/search for transcripts from most agents.

Re: Show HN: Paseo – Beautiful open-source coding agent interface

#54
post #12

"Ship on the go" is so insane to me. It's like people pulling their phones out while taking a piss standing, or having to pull their phones out when the traffic lights are red in a crossing. Just, do one thing at a time, live a life. Shipping code from your phone, whhhyy. Mates, this isn't a flex, it's depression.

I absolutely get that, but I've been working with Claude remote control on my phone as I bicycle every other day. (I have heart and mobility problems and I have to take it gentle with exercise.) I ride to a beautiful spot where there's a bench and sit and use Claude to write features or React components. I'm actually wanting to try this to see if I can rig up a voice-control flow so I can literally just ride my bike all day from park to café to wherever and still be getting code done.

That's the dream, right?

Re: Show HN: Paseo – Beautiful open-source coding agent interface

#56
post #16

I'm the maintainer of Paseo. I didn't submit this, so it was a nice surprise to see it on HN! I'm around if anyone has questions about the project.

Thanks, very nice! Any way to use it with Antigravity CLI/Gemini please?

Re: Show HN: Paseo – Beautiful open-source coding agent interface

#57
Looks nice. The README pitches model-per-task picking but doesn't say much about context management. In coding-agent loops the full system prompt + tool specs re-send on every step — a 30-step task pays the input cost 30x. Prompt-cache headers catch the static prefix, but the per-step diff (file diffs, observation tokens) isn't cached, and that's often most of the input. Auto-summarizing older trajectory into a state vector saved 40-60% input tokens in workloads I've looked at — could be a useful daemon-side concern since users won't reach into each agent's internals.

Re: Show HN: Paseo – Beautiful open-source coding agent interface

#58
post #32
post #28

Earlier quoted context omitted.

very nice and well executed product for what i assume is a small team! (of one?) i guess the obvious question is... if its FOSS and local only, whats the business game plan? you've impressed me that you can ship, but you've basically just matched what eg devin desktop and cursor now do. what next?

Thanks. Yes, team of one right now. I think there's still a lot to do and it feels like I'm only scratching the surface of what's possible in the local layer, so matching what Cursor does is not the end game for me. I want to keep the core open and local-first, the business, when it comes, will be around the convenience/team/enterprise layer. Right now I am letting real usage guide me by paying attention to what team…

good luck sir. if you ever want a job i'm happy to refer you to cognition/devin.

Re: Show HN: Paseo – Beautiful open-source coding agent interface

#60
post #16

I'm the maintainer of Paseo. I didn't submit this, so it was a nice surprise to see it on HN! I'm around if anyone has questions about the project.

Thanks, very nice! Any way to use it with Antigravity CLI/Gemini please?

You can already use Gemini CLI with the ACP adapter, go to the provider settings to enable it.

Antigravity does not yet support ACP, when they do, it will be added to Paseo

Post reply on HN