Standards (or guidelines) on the functioning of Sovrin Agents?


Is there any work being done on standardizing to some degree the way Sovrin Agents should work?


Carlos, yes, in fact I expect this is going to be a focal point of Sovrin development. Per my answer to another question you posted, one of the things I bring to the table is the work of the OASIS XDI Technical Committee on the XDI semantic data interchange protocol. Markus Sabadello, who is also active in the forum, is the co-chair of the XDI TC.

In short, if Sovrin agents use a standard XDI interface, we can establish very broad interoperability without constraining the richness or variety of data that can be shared in the Sovrin ecosystem.

XDI is still young, but you can review the XDI Core 1.0 Working Draft 01 spec to get a good sense for the underlying graph model. We are nearinig completion of a simplified JSON serialization format called JXD that will make XDI much more approachable for developers (it does for XDI what JSON-LD does for RDF).

Happy to discuss further on future Sovrin User Group calls, or to arrange an in-depth call just on this subject.


Thanks Drummond. I’m certainly looking forward for further discussion and work in this area. I think it’s important to allow freedom and creativity for agents, but also to establish a minimal set of parameters to be followed in order to guarantee a level of quality and security in their working and interoperability. This is also in the best interest of the Sovrin community as a whole, being agents a crucial part of the ecosystem.


+1 Carlos. I agree with every word.


We’ve been checking out the current info on XDI, and I must say it looks really powerful. However, I’d be interested in knowing a bit more specifically: On which cases and processes is this represenation used on Sovrin?.. That would give Agencies a better idea on how to use it for interoperability (or even internal agent function).


Hello Carlos,

Right now it’s not used in Sovrin, but we have begun some exploratory work in that direction.
I believe XDI can be useful in various ways, e.g. for data portability, semantic interoperability, as the actual transport protocol between agents and clients, and also for off-chain storage (as discussed in the other thread).

I think it’s likely that there will be some development work to build XDI-based agents on Sovrin, at least I am personally very interested in this.

Of course standards and guidelines for agents will be key topics for the Technical Governance Board of Sovrin Foundation.


+1. I haven’t been active on this thread because I’ve been heads-down preparing for the Rebooting the Web of Trust meeting this week and the Internet Identity Workshop and W3C Verifiable Claims meetings next week (all in San Francisco). But coming out of those should be even more documentation about this entire area, including the use of XDI with Sovrin agents. (Note that I’m likely to be very tied up on those meetings, so please excuse me if I’m not active here on the forum until they are over.)


Thank you very much for your support, guys. I’m indeed looking forward to get more information from you regarding the details of Agent functionality and the role of XDI in it.

I’ve been checking out the documentation on XDI, but I must confess I’m not having an easy time figuring it out, and I’m still not sure where it fits…