# Canister-User threshold signing **URL:** \ **Category:** Developers\ **Created:** [May 7, 2026, 7:20am UTC](https://forum.dfinity.org/t/canister-user-threshold-signing/68427 "2026-05-07T07:20:43Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![MeneseProtocol](https://sea1.discourse-cdn.com/flex023/user_avatar/forum.dfinity.org/meneseprotocol/32/33958_2.png) [@MeneseProtocol](https://forum.dfinity.org/u/MeneseProtocol)\ **Post date:** [May 7, 2026, 7:20am UTC](https://forum.dfinity.org/t/canister-user-threshold-signing/68427/1 "2026-05-07T07:20:43Z")
We thought this was an independent break through on its own so it deserves its own thread. We have been working for sometime on threshold signing in different curves(post quantum ones as well), thanks to the efforts of @MForum Mercatura Forum teams we have been able to create a signing protocol that is between the canister and the user or different canisters and the user. Basically this is relevant for us in terms of purely non custodial holding ie the network can’t act even if compromised on user funds and vis a vis if user is compromised canister can enforce certain policies. This would be extremely ideal if deployed on the new SEV subnets as now there is a risk of node providers going into the canister for the canister’s share. There is a live link to try it out, we welcome anyone who wants to join this effort in reviewing it or providing feedback. It’s a private repo for now, but we intend to bring this infrastructure to community ownership so anyone can use it. Here is a link for the demo it’s basic but obviously you can see this already works very well with the Menese engine we have. The canister runs a presig pool, signing is around 5 seconds total, we can bring it down to 2s in an optimization we are working on. > **[Menese 2-of-2 NIM](https://ohpmj-daaaa-aaaab-qg4wq-cai.icp0.io/)** ---
**Author:** ![quint](https://sea1.discourse-cdn.com/flex023/user_avatar/forum.dfinity.org/quint/32/40449_2.png) [@quint](https://forum.dfinity.org/u/quint)\ **Post date:** [May 7, 2026, 1:29pm UTC](https://forum.dfinity.org/t/canister-user-threshold-signing/68427/2 "2026-05-07T13:29:03Z")
Two things I’m curious about. First, which curves actually work in the demo today versus still being built out? You mention different curves including post-quantum, and I’d love to know where each one is: shipped, prototype, or research. The post-quantum part especially, since threshold PQ signatures are still pretty fresh territory. Second, what’s the timeline for opening the repo and publishing a written spec? For a signing primitive like this, a spec people can review tends to matter more than the code itself, and it’d help anyone wanting to give serious feedback know what they’re actually evaluating. ---
**Author:** ![MeneseProtocol](https://sea1.discourse-cdn.com/flex023/user_avatar/forum.dfinity.org/meneseprotocol/32/33958_2.png) [@MeneseProtocol](https://forum.dfinity.org/u/MeneseProtocol)\ **Post date:** [May 7, 2026, 2:02pm UTC](https://forum.dfinity.org/t/canister-user-threshold-signing/68427/3 "2026-05-07T14:02:04Z")
The current working one with this canister is secp256k1 ECDSA, 1 round signing. What other threshold schemes, Mayo was extensively tested on ICP replica subnet with preprocessing down to 30 seconds, we plan to bring it down to 1-3 seconds window and also on another chain of our’s. Canister also works well with standard ML-DSA and Mayo. Bear in mind all of this could be GPU accelerated at certain levels making it sub second we already published open source on this experiment This repo could be accessible to people from the field for early review, we plan leave it open for some time then put it under community control as a signing service. Probably in the next two months, we have a team that is auditing another project of Mercatura so we will benefit from them auditing the same primitives used > [@quint](#): > > Two things I’m curious about. > > First, which curves actually work in the demo today versus still being built out? You mention different curves including post-quantum, and I’d love to know where each one is: shipped, prototype, or research. The post-quantum part especially, since threshold PQ signatures are still pretty fresh territory. > > Second, what’s the timeline for opening the repo and publishing a written spec? For a signing primitive like this, a spec people can review tends to matter more than the code itself, and it’d help anyone wanting to give serious feedback know what they’re actually evaluating. ---
**Author:** ![MeneseProtocol](https://sea1.discourse-cdn.com/flex023/user_avatar/forum.dfinity.org/meneseprotocol/32/33958_2.png) [@MeneseProtocol](https://forum.dfinity.org/u/MeneseProtocol)\ **Post date:** [June 20, 2026, 3:20pm UTC](https://forum.dfinity.org/t/canister-user-threshold-signing/68427/4 "2026-06-20T15:20:03Z")
Hey everyone, now that the SEV is almost done, we wanted to brainstorm with the community the best way to bring this canister as service under NNS, we envision this to compatible with both hardware wallets so people can build hardware on it(tested very positive results) and/or as a service for those wanting complete autonomy or custody, it also can allow for automated high volume trading without the bottlenecks that could happens to the threshold presig pool and it can scale cheaper. We are going to make a separate topic for it once we test it thoroughly on the SEV, but would really welcome everyone’s feedback on how we can structure it