Were not able to login - because You used github auth without browser interaction. Passkeys are in bitwarden in Chrome...
Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
61–70 of 77 posts
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#62[dead]
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#63The 'record everything on your screen' pitch keeps hitting a privacy wall for teams — even local-only storage triggers a brutal compliance conversation. What's the story for an engineer whose company blocks unaudited desktop recorders by policy? Curious if there's a scoped mode that only captures within a designated app boundary."
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#64Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#65Earlier quoted context omitted.
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 do…
I can, can any vibe coder out there, tho.
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#66Once Screenpipe has months of screen and audio indexed locally, what stops a compromised or over-permissioned agent from querying the whole history through the port 3030 API instead of just the current task's context? Continuous recording solves the memory problem, but it turns "give the agent access to my second brain" into a much bigger blast radius than giving it access to one document. Is there per-query scoping…
We have a permission system that allows you to define rules in the markdown header, such as only allowing a time range to be queried, or modality
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#67I tried this a couple months back. While I found it useful, I also found on my MacBook M1, the fact it was recording all the time, made my CPU melt at high temps all the time. I eventually had to shut it off and stop using because of that. Curious how it runs on more modern hardware? Is it pretty lightweight these days?
Hey sorry about that, we made a lot of work on improving CPU usage, we benchmark it on $200 laptop now and it runs smoothly
You can also configure settings like changing transcription engine or only record meetings or turning off some features to reduce resource usage
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#68Were not able to login - because You used github auth without browser interaction. Passkeys are in bitwarden in Chrome...
Can you expand on this? Usually 2fa allows you multiple options
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#69The 'record everything on your screen' pitch keeps hitting a privacy wall for teams — even local-only storage triggers a brutal compliance conversation. What's the story for an engineer whose company blocks unaudited desktop recorders by policy? Curious if there's a scoped mode that only captures within a designated app boundary."
You can turn off screenshot recording and only capture accessibility data, it still captures most of the meaningful data
Otherwise we usually collaborate with the company to allow screenpipe
Re: Launch HN: Screenpipe (YC S26) – Record how you work and turn that into agents
#70Been a user of Screenpipe to build some "Ai-buddies" for me since last year. Core of that is giving AI a look at what is happening in time space. Without Screenpipe this was rather hard to do at the performance screenpipe gives.
That's interesting, thanks! Indeed we did a lot of work to optimize performance and reliability. I'm curious what does your AI buddies do more specifically and how do you use them?
AI-buddies is now a year old idea and now many other harnesses like Hermes, OpenClaw have most of the same functionality.
Main concept of AI-buddies is to quickly spin up "buddies" who have varying personalities and feel like a real person. Just like real people buddies can talk to each other, share information and have very different functions in your life. For ex a Gym buddy is very different from a "mom" buddy or a Math prof buddy.
Main difference that they are easier to set up for non technical users, and that they can gather input from other real people (for example, if allowed, a Tax buddy can get input from your accountant).