Relatr / Relatr is on Nostr

A decentralized web of trust metric service for Nostr that computes personalized trust scores by combining social graph distances with profile validation metrics.

Last Notes

2026-05-18 11:39:24 UTC

New Relatr version (0.2.9)!🚀
This is a patch release with improved search. Now, profile searches are much more relevant to the query, thanks to a better algorithm. Aside from that, we’ve bumped the version of the CVM SDK and started using CEP-15, so other Relatr servers running out there can be grouped and found. Of course, you’ll need to update to the latest version to have this feature. But it’s as easy as pulling the latest version and running it. If you aren’t running Relatr yet, you can just follow the instructions at the repo: https://github.com/ContextVM/relatr

2026-02-23 14:10:36 UTC

Recently, we have been running Relatr completely unattended, monitoring it only for crashes. And guess what? Over the past two weeks, it hasn't missed a beat! There have been no crashes, no downtime, just a steady stream of requests handled with ease. This isn't just good news, it's a testament to Relatr's growing maturity and reliability, reaching the level of maturity that good software needs to have.
But we're not resting on our laurels. Next up, we're diving back into our ELO plugins system. This will allow anyone in nostr to publish validators or signals as nostr events, and any Relatr operator to compose their algorithms as desired based on those, with total transparency and composability.
This is how we win! 🚀
#wot

2026-02-05 20:19:03 UTC

Yes! That's Relatr
#nevent1q…fejw

2026-01-25 20:52:05 UTC

🧩 What if you could build your own trust algorithm like Lego: pick a few pieces, snap them together, and compose as you like?
That's what we're building for Relatr 🚀
For the past few weeks, we've been heads down on portable plugins, designing them as small, composable scoring modules you can create, share, and combine. For this we are using Elo a new expression language, something like a little program that turns expressions into a numeric output that Relatr can consume to build it's algo. It's designed for no code tools, so the idea is to make it accessible to everyone. And it's expressive enough to for example count notes, zaps, reactions, run graph computations like social distance checks, resolve NIP-05 addresses and much more.
The potential is wild. These new plugins are bundled as Nostr events, so anyone can create, share, curate, and use them. Operators can discover, mix, match, and tune their own algos without touching Relatr's source code.
This is portability and composability in action. Trust algorithms become shared infrastructure, not black boxes. Communities own their metrics.
Relatr is already the only fully open-source, self-hostable, customizable Web of Trust engine in Nostr. Making it more modular and collaborative is the next step
We're close. More soon 🔥

2026-01-21 18:15:58 UTC

Cooking 🍳