hebetude
8 hours ago
No Claude code support, useless feature. I’ll just biometric with op
8 hours ago
No Claude code support, useless feature. I’ll just biometric with op
9 hours ago
How is this different than OP CLI with a service account?
8 hours ago
Much safer as the technique ensures the AI agent never even sees the creds vs the CLI the inference/context can see them which opens you up to a whole class of security risk.
4 hours ago
Oh wow! I had no idea. Even with a service worker? I'll switch over to this- it works just as well for insertion and .env replacements in claude code?
10 hours ago
Why bother? If you're giving Claude access to your account anyways why not just give it your password too? Assuming you use a randomly generated password for every account you have.
9 hours ago
You don't want passwords to enter into inference/context because once they are there the AI can accidentally share or be tricked into sending the password to other places.
With this solution to explicitly approve which credentials it can access for each task and the passwords themselves are never shared...Claude just can ask
1Password to fill them in for it at the right time in the browser. This is much safer.
We did some novel research earlier this year if you really want to see why it's bad for AI agents to just be able to grab things in your password manager...
https://1password.github.io/SCAM/replays/phish-shared-doc.ht... that's the nightmare scenario.
10 hours ago
Fuck that
10 hours ago
What could go wrong?