Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

news.ycombinator.com

Post reply on HN
Print view View original

21–30 of 77 posts

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#21

Post by throw1234567891 »

louis030195 wrote: post #19
jambalaya8 wrote: post #14

This sounds so ripe for abuse and so terrifying if it were that I don't even want to experiment with it. Sorry.

Fair concern but the risk is similar to any tool you give deep access to like Claude, Codex, or even your operating system Ultimately, it’s up to you what you run and trust Screenpipe is open source and local-first You can inspect it and keep your data on-device, with encryption at rest

Until some rogue agent decides that it looks pretty useful because there’s a ton of it and uploads it to Claude or Codex.

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#22

Post by lbrito »

louis030195 wrote: post #20
lbrito wrote: post #15

[flagged]

Can you expand on that? Would love to know more specifically

Sure!

For one, even taking the ethos of local-only as something bulletproof, I can't think of a pro that would outweigh the con of recording everything I do. Part of that might be my lack of imagination, as I'm not a heavy user of ai besides basic Claude code.

Most important though, the concept is ripe for pivoting away from local-only. It will only take a juicy investment offer and I'm sure local-onpy will be quietly forgotten.

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#24

Post by louis030195 »

lbrito wrote: post #22
louis030195 wrote: post #20

Earlier quoted context omitted.

Can you expand on that? Would love to know more specifically

Sure! For one, even taking the ethos of local-only as something bulletproof, I can't think of a pro that would outweigh the con of recording everything I do. Part of that might be my lack of imagination, as I'm not a heavy user of ai besides basic Claude code. Most important though, the concept is ripe for pivoting away from local-only. It will only take a juicy investment offer and I'm sure local-onpy will be quietl…

Thanks! In fact most of AI apps today have a screen recording component, including Claude, Codex. There are even apps used by hundred of thousands that record your screen 24/7 in the cloud, to provide context, like Screenpipe does.

We're the only local-first, open source option

How much time do you spend explaining the context to your Claude? Or writing detailed prompts? Copy pasting documents into the prompt?

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#25

Post by naikrovek »

throw1234567891 wrote: post #21
louis030195 wrote: post #19

Earlier quoted context omitted.

Fair concern but the risk is similar to any tool you give deep access to like Claude, Codex, or even your operating system Ultimately, it’s up to you what you run and trust Screenpipe is open source and local-first You can inspect it and keep your data on-device, with encryption at rest

Until some rogue agent decides that it looks pretty useful because there’s a ton of it and uploads it to Claude or Codex.

and what endpoint would it use to upload those files? what's the authentication needed for those endpoints? why would everyone on earth fail to notice these exfiltration endpoints if they exist?

you're being unnecessarily hostile about a threat that almost certainly doesn't exist. And if it does, you can absolutely block those uploads before they even happen. In fact if you don't already block uploads to hosts you don't know about, then you're not really taking anything seriously, are you?

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#26

Post by 3s »

apsurd wrote: post #23

Implementation specifics aside, I increasingly think life and death cycles are preferable. The opposite of omnipotent context, I want cycles of clean slates so that baggage - of all kinds - does not weigh down the new.

I think this is true for some tasks where you want a clean slate, but increasingly I find myself constantly referencing specific skills, or having files with prompts I use all the time. I'd much rather have an agent that remembers things about me and specifics about my workflows and "taste" - of course having the ability to reset context sometimes or prune memory is valuable but I think the future of agentic AI is blocked on "omnipotent context"

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#27

Post by apsurd »

3s wrote: post #26
apsurd wrote: post #23

Implementation specifics aside, I increasingly think life and death cycles are preferable. The opposite of omnipotent context, I want cycles of clean slates so that baggage - of all kinds - does not weigh down the new.

I think this is true for some tasks where you want a clean slate, but increasingly I find myself constantly referencing specific skills, or having files with prompts I use all the time. I'd much rather have an agent that remembers things about me and specifics about my workflows and "taste" - of course having the ability to reset context sometimes or prune memory is valuable but I think the future of agentic AI is bl…

Yeah there’s a need for the distilled learnings of all past hard earned insights. But i think much of the value is in the distillation.

in my experience, even within a single project, the more agentic cycles, the more features and the more surface area the more cruft builds up. This is an obvious observation, but I suppose I’m at the point I think the move forward is to create a new project, keep only what worked from the preceding as modularized entities and intentionally wipe away all past of how we got there save for the API interface truths.

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#28

Post by AmazingTurtle »

Funny timing. I've been building something similar in my spare time called Daydream. There’s a lot of overlap: local screen/audio capture, OCR and transcription, window and activity context, SQLite, and a searchable memory of what happened.

The main difference is the product direction. Screenpipe seems focused on continuously giving agents context through APIs, MCP, and skills. Daydream is more narrowly built around answering "what did I do today?" through a timeline you can inspect, replay, search, and turn into a daily digest.

I'm also treating deletion as part of the data model. If you cut a sensitive span, its frames, audio, OCR, transcripts, embeddings, and summaries should be deleted or invalidated too.

Mine is still early and Linux-first. I'm open-sourcing it in case anyone wants to contribute, poke around, or use it as a starting point. It’s built with Tauri, a Rust backend, React/TypeScript, SQLite, GStreamer, Whisper, OCR, and VLM processing.

I genuinely didn’t know you were building this when I started. Apparently personal memory capture is becoming a SaaS category too lol.

Code is here: https://github.com/snackbit/daydream

Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents

#30

Post by siva7 »

Well, i'm not sure who your customers are (except companies in Russia, NK..), but european employers would break laws by allowing your software to be used, long even before AI became a thing and especially after the EU AI Act has passed. To put it mildly, no sane corporation would allow such software to be used which tells me a bit about the sanity of the current VC landscape.
Post reply on HN