<unnamed>

Public Key

npub1xdtducdnjerex88gkg2qk2atsdlqsyxqaag4h05jmcpyspqt30wscmntxy

Profile Code

nprofile1qqsrx4k7vxeev3unrn5ty9qt9w4cxlsgzrqw752mh6fduqjgqs9chhgpzemhxue69uhhyetvv9ujuurjd9kkzmpwdejhgm08n70

Publishing to

relay.primal.net

Last Notes

2025-09-09 19:05:41 CEST

- reply

I mean if keys can't be extracted and operator just kills the enclave then keys are gone and LN node's channels are force closed and node's sats are gone with the keys and operator gets no benefit from the shutdown. Or am I missing something?

2025-09-09 18:47:56 CEST

- reply

My point is the motivation for taking your funds for myself is obvious, while taking them to no-one's benefit - not so obvious.

2025-09-09 18:28:14 CEST

- reply

It's not a rug, funds aren't stolen, but destroyed - enclave is terminated and keys are lost, a very different set of incentives to do this.

2025-09-09 18:12:45 CEST

- reply

I think the difference btw "we are 100% sure this mint runs this code" vs "we don't really know what it's doing but some checks make it look fine" is significant.

2025-09-09 17:43:44 CEST

- reply

I would separate rug pull risk from service failure risk. Cashu is custodial without unilateral exit so with or without enclave, if mint goes down, funds are inaccessible. To me it looks like rug pull risks are minimized with enclaves - we could get a lot of independent mints all running verified code, which sounds quite awesome. The rest are reliability questions: backup might be sync real time, coordinator can live in another enclave, dns record can be controlled by the current master, updates don't require a blockchain - just a set of signatures by pre-declared maintainers (that's my current approach). Not that it's a proven set of solutions, or that I am against the blockchain, but I think more experiments are needed to figure out the actual problems and solutions.

2025-08-07 15:05:04 CEST

Clone yourself https://github.com/nostrband/askexperts

2025-07-21 09:09:09 CEST

- reply

Hi, it should have been fixed, does it update now? Apologies.

2025-06-24 09:02:23 CEST

- reply

I am the creator, there were no commits there for a while, I didn't remove anything.

2025-06-23 12:43:16 CEST

- reply

Try reloading the page, etc. We changed nothing there, something might have changed with your extension/browser setup.

2025-05-29 18:00:52 CEST

- reply

After you go through hell of getting your android and ios dev accounts

2025-05-29 17:59:52 CEST

- reply

Found it, says Connecting and then Request timeout.

2025-05-29 17:54:02 CEST

- reply

How can I connect the nwc wallet?

2025-05-26 16:17:02 CEST

- reply

Awesome! Apologies for the issue!

2025-05-26 10:34:56 CEST

- reply

Fixed, could you please open use.nsec.app then wait a bit, click Reload button at the header when it appears to update it, then retry. Thanks you very much for this feedback!

2025-05-24 07:26:18 CEST

- reply

Just use

2025-05-23 23:17:19 CEST

- reply

That's helpful! Will get back to you

2025-05-23 23:16:04 CEST

- reply

VS plugin https://github.com/RooCodeInc/Roo-Code

2025-05-23 23:14:53 CEST

- reply

I felt the same way with main projects, but starting from scratch is much easier, just make me a prototype and we'll figure it out. There are enough important pieces to think about and leave mundane to the bot

2025-05-23 21:24:16 CEST

- reply

I've been vibing a side thing in the background for a week. I break it into small tasks, think through the architecture and major important pieces. Then roocode and Claude do next task in several minutes. Sometimes it goes crazy and I had to rollback. Sometimes it's stuck and I have to figure things out. But if you treat it like a super productive junior, it's fascinating. It would take me at least 3 weeks full-time to get where we are now. Just don't think it can do big things alone.

2025-05-23 21:19:10 CEST

- reply

Where are the other 86?

2025-05-23 09:36:31 CEST

- reply

My apologies. Could you please try again now?

2025-05-22 08:53:54 CEST

- reply

What's the problem with nsec.app ?

2025-05-21 11:07:24 CEST

- reply

It has balance limit of 1000 sats, it's in testing-only mode right now.

2025-05-20 15:32:29 CEST

- reply

Well while we were talking RooCode looked through one third-party golang codebase, figured out how to do what I need, and wrote me a ts function to solve my problem. I tried doing that by hand before and just gave up. Sometimes it's super helpful.

2025-05-20 13:36:44 CEST

- reply

I'm only playing with it atm, no good number yet.

2025-05-20 09:35:50 CEST

- reply

Should Agent Cards also be an addressable nostr event? Either as a copy of the json file, or as some summary just to allow the discovery of json url using hashtags, wot etc. I'd prefer the former - some agents might very well be behind NAT and shouldn't be forced to need an http server.

2025-05-20 08:57:21 CEST

- reply

Sorry I don't see any other way to pass ln address with NWC - lud16 is included now, maybe that's something on your side? Are you setting the NWC string into an existing profile that already had ln address? Maybe Freerse uses address from NWC only if profile didn't have one already?
Yes it's using relay.zap.land

2025-05-19 18:29:41 CEST

- reply

Does the NWC string that you're getting now include &lud16=<address> ?
Yes works great w/ @nprofile…q7y6 !

2025-05-19 16:26:15 CEST

- reply

Ah ok I found it, added lud16 to NWC now - could you please try again checking if LN address will be taken from NWC?
I recommend managing the wallet using Alby Go or similar app for now, I don't have immediate plans to add transfers etc to zap.land

2025-05-19 15:36:16 CEST

- reply

Hi, this isn't really a wallet app, it's just here for you to generate NWC string and start using it elsewhere.
Zap.land gave you the ln address which you could put into your profile, is that what happened? NWC isn't really related to ln address, these are separate things

2025-05-17 09:42:21 CEST

- reply

Ok see it below now.

2025-05-17 09:41:30 CEST

- reply

This is awesome! Any working prototype out these?

2025-05-15 11:59:14 CEST

- reply

This was basically my first shot, I just used cursor with claude 3.5 sonnet. What do you use?

2025-05-15 07:12:10 CEST

- reply

Although I am looking at the diagram, the LN payment itself is offline, but looks like invoice generation still needs receiver to be online, which means something like zaps isn't possible offline.

2025-05-15 06:54:37 CEST

- reply

Awesome, thanks! Offline receive is a great capability.

2025-05-14 17:56:12 CEST

- reply

Don't worry, only llms will read it

2025-05-14 17:54:57 CEST

- reply

How so?

2025-05-14 16:17:50 CEST

- reply

Automatic wallet creation by apps for new users is one of the goals of the nwc-enclaved, and it supports zaps.
The tradeoffs with Spark are that it's a new tech and quite complex and not many people can understand it. And Spark operators have to be trusted, although I'd say they should run in enclaves too. And no fees for now but they will come and hopefully will be competitive. And 'unilateral exist' is 'coming soon' at the moment. And your wallet still has to be online 24/7 to receive payments if I understand it correctly. So - pick your poison. I think it's very promising, but wouldn't bet 100% on it right now.

2025-05-14 15:47:17 CEST

- reply

Could you reload and retry please? Works fine here

2025-05-14 15:46:47 CEST

- reply

We will create spank-based on too, it's a different set of tradeoffs.

2025-05-14 15:34:13 CEST

Lightning wallet in 1 click!
I vibe-coded a trivial wallet-creation app today: https://zap.land
One click and you get NWC string that you can put into Alby Go to topup the wallet and send payments, or put into Damus/Amethyst to send zaps. You also get LN address that you can paste into your Nostr profile and receive zaps to this wallet.
https://image.nostr.build/9ec71b3e9425129120863d812f9ece3f671838757b4a20ef34e2a840ce64b471.jpg
The backend of this is a development instance of a new custodial lightning wallet deployed in a trusted execution environment. That means you get better security and privacy guarantees, along with great usability. It's open source, you will be able to launch an instance of it yourself and clients could discover your instance along with ours and let users choose. The source code is here: https://github.com/nostrband/nwc-enclaved
This is an invitation for public testing and for your feedback. The wallets created have the max balance limit of 1000 sats and will likely be turned off at some point when testing is over and dev instance needs to be shut down. This is not the real thing, it's a toy to get a taste of it!
The service is based on "phoenixd" wallet with fully automatic liquidity management, which means - fees. Phoenix charges 4 sats + 0.4% per payment - we charge extra 1 sat, phoenix charges 1% plus mining fees for liquidity - we add nothing here, we also charge 1 sat per day to avoid getting spammed. When I say "we" I mean "the code inside the enclave" - it will only use revenue to pay for itself inside the enclave, not to make a profit.
Anyways, send some feedback!

2025-05-14 12:56:25 CEST

- reply

That's what I'm trying to make Cursor do and it's failing miserably :(

2025-05-12 09:21:15 CEST

- reply

Hello, no nsec.app doesn't interfere with your nip05 set in your profile - we assign you our own nip05 but you'd need to enter it into your profile manually for other people to see it.

2025-05-02 12:36:22 CEST

- reply

I could give it but I recommend trying something else, strfry has high internal latency for delivering new events to active reqs (>300ms). I plan to switch from strfry myself, here is a nip46-specific khatru-based relay by @npub1h49…9kay https://github.com/dezh-tech/ddsr/tree/main/bunklay - haven't tried yet, but looks promising.

2025-05-01 13:40:52 CEST

- reply

nsec.app is running over relay.nsec.app which is a normal strfry. It's actually quote slow for ephemeral events processing - it has internal latency of at least 300ms for delivery of new events to active SUBs, plan to switching to something more specialized. Aside from that, again, try the secure enclave thing in settings.

2025-04-30 09:29:42 CEST

- reply

relay.nostr.band now let's you search through these events

2025-04-27 16:43:21 CEST

- reply

In our case app in tee generates a key to be used for encrypted comms and that pubkey is included in the attestation

2025-04-27 16:42:22 CEST

- reply

Comms with enclave are e2ee that's critical. Attestation of nitro can have pubkey, nonce or small piece of data

2025-04-26 09:03:11 CEST

- reply

They should have applied for a grant to pay influencers, isn't that how it's supposed to work?

2025-04-25 22:52:30 CEST

- reply

Many people seem to have worked on this, why don't you mention any actual project?

2025-04-25 20:40:50 CEST

- reply

You could make search requests to relay.nostr.band and other search relays it's nip50

2025-04-24 14:49:11 CEST

- reply

kinostr

2025-04-24 14:43:21 CEST

- reply

Ah also you app is passing zap request in base64 encoded form, instead it should be url-encoded.

2025-04-24 10:08:27 CEST

- reply

Yes, add 'relays' to 9734

2025-04-24 10:07:02 CEST

- reply

NIP is fine, your app just needs to add 'relays' tag to 9734

2025-04-24 10:02:55 CEST

- reply

I see, sending you a signal DM please check

2025-04-24 10:02:10 CEST

- reply

I see zap receipts don't work because zap request 9734 event doesn't have "relays" tag (https://github.com/nostr-protocol/nips/blob/master/57.md#appendix-a-zap-request-event) - zap receiver doesn't know where to publish the receipt.

2025-04-24 09:59:04 CEST

- reply

Ah yes! Didn't notice the Zap symbol on kinostr. Did you try to figure out how to add some wallet integration? For one-click zaps or for creating wallets for signing-up users?

2025-04-24 09:52:45 CEST

- reply

Cool! Which of your apps have zaps? Zap receipts are published by receiving wallet, so I don't think your app is responsible for that.

2025-04-24 09:33:44 CEST

- reply

How about ln/zaps?

2025-04-24 09:13:50 CEST

- reply

Did it work?
Btw did you try integrating LN/zaps to your vibed apps?

2025-04-23 13:45:19 CEST

- reply

Could this README work? https://github.com/nostrband/nostr-login/blob/main/README.md

2025-04-23 11:33:34 CEST

- reply

Have you tried asking it to add nostr-login library as a drop-in solution?

2025-04-23 11:25:32 CEST

- reply

Yes we use Ghost themes but not Ghost itself. It can be self-hosted but there are currently some drawbacks when doing that (no server-side rendering), instructions here https://blog.npub.pro/post/how-to-self-host-on-github/

2025-04-12 10:10:58 CEST

- reply

Thanks for the link, good read!

2025-04-11 21:53:27 CEST

- reply

Amber works great for apps on the same device. Cross device it has same issues as nsec.app - Amber is killed by mobile OS and can't process requests reliably.
This enclaved signer can accept your nsec and process requests to it. You still retain your nsec and can control the signer with it, that's what nsec.app does. Your local key just can't be always-online.
Does it make sense?

2025-04-11 21:41:21 CEST

- reply

Nothing to prove here, if you trust AWS virtualization infrastructure, then use it, otherwise ignore. There is always some root of trust.

2025-04-11 21:22:02 CEST

- reply

Why?

2025-04-11 21:03:59 CEST

- reply

If the code is well written, enclave operator can only turn it off, not mess with it in any other way.

2025-04-11 21:02:49 CEST

- reply

Sounds fine, will see what I can do. One issue at this point is that if enclave is restarted, all the state is gone, so storage of long term assets is a challenge, especially while it's not yet production ready. But we will solve that, so eventually your idea is gonna work.

2025-04-11 20:57:13 CEST

- reply

Ah got it, you pass the bunker url to creator, and they try to export and that only succeeds once. Sounds good!

2025-04-11 15:20:05 CEST

- reply

So you create an nsec for Taylor Swift and start reposting her insta pics on nostr, those get zapped, eventually you tell her about those sats and invite to take over the nsec. Is that the set up? Wouldn't you want to prove to enclave, somehow, that it's her? Otherwise someone could front-run her?

2025-04-11 14:53:43 CEST

Store your nsec in a secure enclave!
We've spent a big part of 2024 trying to make a reliable non-custodial signer with nsec.app. It hasn't worked out perfectly - remote access to keys stored on a mobile device is still unreliable, especially on iOS.
That's why we got very interested in AWS Nitro Enclaves - h/t @nprofile…j64l and MapleAI team for inspiration!
The idea is to have an open-source custodial signer and deploy it in an isolated environment that provides attestation for the deployed code. Anyone would be able to reproduce the code build and verify that the signer is running the correct code in a safe environment.
So here it is: https://github.com/nostrband/noauth-enclaved
Each instance of a signer deployed in an enclave announces itself on Nostr. We added some Nostr to attestation provided by AWS - the report is linked to pubkeys of a person building the code image, and a person running it. This way you can choose a signer based on your own preferences.
It's already integrated into nsec.app, go to Settings => Secure enclave and try to upload your keys to the server. The signer API is described in README.
https://image.nostr.build/c94c04248420e49a0c9c3c13727943d47b629ec63c9a5078fd9302f675e7032f.png
DO NOT deploy your real keys just yet - it's running in a production environment, but the code hasn't been well audited yet. First we need some community members to review the code and reproduce the code image.
The signer can also be used to generate throwaway keys for automated testing of nip46 implementations, check out "generate_test_key" method.
Looking forward to your feedback!
(Signed by my nsec from the enclave)

2025-04-09 21:18:05 CEST

- reply

Even Protocol Enhancements section doesn't mention payments, they probably hope to get everyone on their closed platform eventually. Same with discovery.

2025-04-09 21:11:44 CEST

- reply

That's what I was trying to say

2025-04-09 21:03:57 CEST

- reply

Exactly, don't see payments mentioned which makes no sense

2025-04-01 18:17:20 CEST

- reply

Fixed

2025-04-01 13:18:37 CEST

- reply

No, you can only change contributors. Admin can change settings, contributors post content.

2025-04-01 11:50:00 CEST

- reply

Hi, we can talk here or over nip17 dms

2025-03-31 15:09:27 CEST

- reply

Retention graph is obviously broken, will look into it.

2025-03-30 10:30:00 CEST

- reply

Congrats on good progress!

2025-03-28 12:58:18 CET

@nprofile…vu4p are there api docs for datamachine?

2025-03-28 11:44:24 CET

- reply

I'd be happy to use NWC as your docs recommend but I don't see how I could create the sub-wallet (nwc string?) without using an API.

2025-03-26 19:12:08 CET

- reply

I ignore that too for now bcs kind 1 is the only source of comments on most sites now, without them all comments sections will be empty.

2025-03-26 19:10:50 CET

- reply

Nice!

2025-03-25 20:26:50 CET

- reply

Alby should just queue requests that require a prompt and open them one by one.
Asking what is pre authorised is not great bcs signer might dynamically decide if the prompt is needed i.e. if client is trying to erase the contact list etc

2025-03-25 12:48:42 CET

- reply

Please tell me more about the nightmare part? What wasn't working or was confusing? Thanks!

2025-03-25 07:44:53 CET

- reply

Aside from only 24133 kind, what other benefits do you see with it?
Does it store these ephemeral events for awhile? We need that for nsec.app
Auth to read would break nsec.app bcs it wouldn't be able to monitor unanswered requests to send push notifications.
Focus on lowering latency as much as possible. Maybe ignore any req without kind and authors to reduce any irrelevant load.
Focus on stability, most bunkers use a single relay and if it's stuck then users are furious.

2025-03-24 20:49:55 CET

- reply

Yes try adding this to Code injections in settings. Will need a bit of a guide though

2025-03-24 20:39:38 CET

- reply

Can't use without nip46 😭

2025-03-24 20:36:53 CET

- reply

Cool!
#nevent1q…h7tj

2025-03-24 20:27:39 CET

- reply

👀
#nevent1q…08vh