Cashu R&D

Is there any open source code to self host this or a builderlab equivalent?

Yes its open source. https://github.com/block/buzz

in case u need a dev break, day 2 of btc++ in toronto is streaming now https://x.com/btcplusplus/status/2083244093686353960

Hi @callebtc, are there specific tools, prompts or AI skills you recommend for security and vulnerability testing? I'm looking to equip my AI agent to run automated security checks on my code.

Good catch mate 👍 thank you

Hi guys, I recently built the Routstr app on Alby Hub and I'm looking for user feedback. I would love for you all to test it out when you have a moment. Branch: https://github.com/welliv/alby-hub/tree/routstr Demo: https://primal.net/e/nevent1qqs85a2ruvpvjdeq5vg9sagefznfqk55l3hg8cr5f8qmapjdvpp4lyqjwerrg

Ive built a fundraiser for agents based on roustr if anyone is interested, I dropped the post here but it seems it was removed ..

Hi, I am Diego from Plebeian Market. We’re investigating a wallet-loss case from an older Cashu client/mint flow. Lightning payment succeeded and the mint reports that the mint quote reached the legacy ISSUED state. The wallet briefly showed the minted balance, but after refresh the proofs were gone. We’ve found no persisted wallet proofs, no recoverable NIP-60 token event, and the client appears to have generated the mint outputs non-deterministically. NUT-09 normally requires the original BlindedMessages, which we no longer have. Before concluding the ecash is unrecoverable, does Nutshell (or older Nutshell versions using the legacy ISSUED quote state) persist any additional quote-linked minting material that could allow an operator to identify or safely recover the exact blinded outputs/signatures for that issuance? Or, without the original BlindedMessages / deterministic wallet recovery material, is NUT-09 recovery fundamentally impossible?

If the outputs were generated non-deterministically, there is little to be done. What wallet was this?

It was Plebeian Market’s NIP-60 wallet, using @nostr-dev-kit/wallet 1.0.0 with cashu-ts 2.9.0 I traced the deposit path: NDK created the CashuWallet without a bip39seed, and mintProofs() was called without a counter or custom output data. In cashu-ts that falls back to OutputData.createRandomData(), so the mint outputs were indeed non-deterministic The Lightning payment settled and the mint completed issuance, but the client lost the resulting proofs before a durable NIP-60 7375 was available If useful I can share the relevant code paths/versions here

Yea there have been problems with NIP-60 wallets. If the proofs were not persisted and were not generated deterministically, there is nothing to be done.

Thanks, that confirms what we were seeing in the implementation. We’re checking the affected wallets individually for any surviving NIP-60 proof events before treating them as unrecoverable

🥜🥜 Cashu dev call 36 this Thursday August 27th 16:00 UTC 🥜🥜 Timezone converter - PLEASE double-check your local time. 👉 Meeting link Draft Agenda Note: We will not record these meetings and Chatham House Rule applies.

Fam https://github.com/cashubtc/BTCNutServer is not working with the latest btc pay server version

There was a vulnerability discovered so we disabled the plugin so instances can't be exploited. Someone from the team was going to contact @d4rp4t or @callebtc but I guess that never materialized.

starting in 15m. join us if you can!

Fully re-designed cashumints.space, now you can also review offline/rugged mints! 🎉 https://cashumints.space