From an author:
I wanted to share crypt.fyi - a free, open-source tool I built for securely sharing sensitive data/files. It uses client-side encryption and zero-knowledge architecture.
Key features:
- Zero-knowledge architecture
- End-to-end encryption using AES-256-GCM (actively investigating post-quantum encryption options)
- Self-hostable
- Suite of configurations (password, burn after read, max read count, ip/cidr-allow list, webhooks)
- Strict rate-limiting
- Strict CSP to mitigate supply chain attacks
- Web, cli, and chrome-extension clients
- Fully open source (Github)
The problems I aimed to solve: Many people share sensitive info (passwords, keys, etc.) through email, Slack, or SMS - which often leaves plaintext copies in multiple places. Existing solutions either require accounts, aren’t open source, or have security/privacy/ui/ux/feature/config gaps/limitations.
crypt.fyi is built with privacy-first principles:
- No logging of sensitive data
- No analytics or tracking
- Separation of web and api servers
- All encryption/decryption happens client-side using shared cross-platform cryptography primitives from noble cryptography
- TLS encryption for all traffic
- Encrypted data is automatically destroyed after being read with strong guarantees around once-only reads
The entire codebase is open source and available for review. I’d love to get feedback from the privacy community on how to make it even better!
Interesting!
How do I decrypt the data when I download my friends holiday photos he shared with me?
How big can they be, can I share my favourite Judas Priest album with my niece?
This is not a criticism, but was there some specific reason for using GCM?
Good work!
How do I decrypt the data when I download my friends holiday photos he shared with me?
They’re decrypted automatically in your browser via the
keyin the URL and additionally a password (assuming one was set when created). Both thekeyand password are used to encrypt the contents so the key alone is not sufficient to decrypt the contents. Regardless, it happens automatically entirely in your browser without ever sending the key or password to the API server.How big can they be, can I share my favourite Judas Priest album with my niece?
I have the limit set to ~500kb right now. That’s after encrypting the contents. How large is your favorite Judas Priest album? Maybe I can uptick to accommodate it haha.
specific reason for using GCM
Given the different tradeoffs on performance, security, and implementation complexity, GCM seemed like a reasonable choice. I’m making sure to use the OWASP recommended PBKDF iterations 1 2. I’m also looking into post-quantum options recommended by NIST 1.
Edit: Hi & welcome! Nice to come by and discuss these kind of things!
How is the key circulated? Over RSA or something? Or do you have to send the link+key somehow to the recipient?
Yeah GCM is nice with the inbuilt authentication. AES 256 I guess?
NIST is aaaabout to chose an algo, right? I dug deep down in all that quantum stuff like a year ago, but it didn’t seem like they’d chosen Rivests algo just yet ^^
BTW is the GCM adding a lot of space? I’m on AES CTR (which just aligns to the block size) + RSA for authentication.
The key is transmitted in a URL query parameter. I’m planning to optionally have it transmitted separately from the URL but ultimately, the decryption key would be transmitted via otherwise insecure / normal means. This is where the understandable and healthy critique around the security/privacy of the tool stems. I shared with another user that this tool is an incremental step in the direction of more secure and ephemeral transmission of data with convenience and accessibility as a core tenant of the tools existence. Yup it is AES 256 and I believe NIST has finalized post-quantum recommendations. I’ll likely be using ML-KEM which increases the resulting data size considerably but is also considerably faster.
Thanks for the explanation!
Yeah, you could let the user encrypt the sensitive data with “your” public key for example (or use ssh I guess). For sharing keys, that’s more complicated of course, especially if the person that is going to get the key doesn’t know it will get the key beforehand (or the key can be encrypted with their shared public key).
Also note that there’s OnionShare, too. You don’t need a TLS certificate or domain, don’t need to port forward and can run it from home safely, routes over Tor so very hard to know you are even sharing something, well known and open source etc.
This detracts a bit from the convenience factor in that it requires both sender and receiver to download an app as opposed to having access to a browser. It is however much safer as the client is static and versioned so the privacy guarantees are better.
In the same way that having a hammer requires a second tool on-hand when I can otherwise just bang the screws into concrete with my adjustable wrench?
- Encrypt
Your secret is encrypted right in your browser before it ever leaves your device. Only people with the special link, that you’ve explicitly shared, can decrypt it.
- Share
Share the secure link with your intended recipient. The link contains everything needed to decrypt the message, unless a password is specified.
So you still need to share a secret - the URL - over some other secure channel, otherwise you still risk leaking your secret data. Sure, it’s easier to share a short piece of text than a large file, but this still doesn’t solve the motivating problem that the tool is setting out to solve, namely:
Why crypt.fyi?
The Problem with Common Practices
Every day, sensitive information like passwords, API keys, and private data is shared through insecure channels:
Is it too hard to ask normal people to use asymmetric cryptography?
A web-based tool for temporary data sharing like this could make public-key crypto easy for non-technical people by doing all of the key management for the user. “Your key has been generated and stored in your browser. Share this public key with your correspondent and they can send secure messages to you.”
The one sticking point is that your recipient needs to visit the site before you can send your vacation photos to them, but is it really that hard?
Creator of crypt.fyi here! Thank you for pointing out the lack of clarity in the ‘why’ statement. It really should state that sensitive data is shared through insecure channels in plain text. The in plain text part is key, as it leaves the contents indefinitely vulnerable.
Is it too hard to ask normal people to use asymmetric cryptography?
Peoples eyes have glazed over at the mere sound of the words ‘asymmetric encryption’. It’s a bit out of touch of a statement to make haha. I believe these individuals still deserve more accessible tools that make them incrementally more secure and maybe it can be a gateway to even more secure/private solutions.
Nice to see you here!
