Atlas Protocol
Atlas vs Nostr
Nostr proved that cryptographic keys and dumb relays are enough to bootstrap censorship-resistant publishing. Atlas builds on that minimalism with durable identity, typed records, and protocol-governed discovery instead of relay-by-relay rules.
A simple protocol for signed social publishing
Nostr keeps the core model refreshingly small. Users sign events with keys, clients send them to relays, and other clients subscribe using filters. That makes publishing more portable than on a traditional platform because your account is not tied to one company database.
That simplicity is real power. NIP-01 gives the protocol one event shape, relay subscriptions, and replaceable or addressable events for some kinds of data. Nostr works well as an open conversation layer because it stays easy to implement and hard for one company to own.
Simple relays mean more work above them
If you want a full application protocol, you eventually need more than signed events over relays. You need stronger answers for identity, permissions, structured querying, trust, moderation, governance, and persistence.
Nostr can support many of these through NIPs and app conventions, but much of the burden still sits with clients, relays, and communities. That openness is flexible, but it also means important behavior fragments.
In practice, this also means Nostr is less decentralized than it first appears. Most clients subscribe to a small set of relays and expect that to represent the network. If a public figure were banned or quietly suppressed by the biggest public relays, their reach could collapse overnight.
Fiatjaf showed this problem directly by publishing identical notes to his own relay versus popular public relays and seeing a dramatic difference in responses. The missing piece is a more neutral, semi-automatic discovery and gossip layer. Atlas is designed to add more of that layer, so reach depends less on a handful of dominant relay operators.
- Identity: public keys are a strong base, but human-friendly naming and higher-trust identity stay layered on later.
- Permissions: remote signing exists in NIP-46, but it is still an optional extension rather than one shared app permission model.
- Data and querying: the event model is flexible, but complex apps still need extra indexing and interpretation on top.
- Trust and discovery: clients, relays, and local communities still decide what to surface, mute, or believe.
That is the tradeoff: Nostr is elegantly minimal, but Atlas is trying to specify more of the difficult middle layer that apps keep reinventing.
Atlas treats the network as shared application infrastructure: stable identity, typed envelopes, registries, trust, governance, and incentives sit closer to the protocol.
Atlas gives identity more structure
Identity + PermissionsNostr starts with a keypair, which is a strong base. But user-friendly naming and stronger identity assurances are added later. NIP-05 maps keys to DNS-based identifiers, and remote signing in NIP-46 is optional.
Atlas aims for a stronger default: durable protocol identity, safer custody for the root key, delegated app permissions, and identity that does not depend on DNS aliases to feel native.
Naming and signing workflows can be improved, but they are not one built-in identity model.
Apps can receive limited permission without permanently owning the most important secret.
Atlas organizes data for apps, not just event streams
Structured DataNostr's single event format is elegant and versatile. NIP-01 even defines replaceable and addressable events for some kinds of data. But relays still mostly store and forward generic events, so complex apps often need extra indexing and interpretation.
Atlas uses typed envelopes, validators, and specialized registries so records can be served more like structured database entries than generic social events.
Flexible and portable, but meaning and indexing stay more scattered across apps and relays.
Designed for validation, querying, and serving structured app data at speed.
Trust, moderation, and governance are less ad hoc
Trust + GovernanceIn Nostr, relays and clients decide filtering, access, ranking, and moderation. That can be resilient, but it also pushes hard social questions to the edges.
Atlas brings trust and governance closer to the protocol itself, so discovery, reputation, and rule changes do not live only in scattered client heuristics or private relay policies.
Atlas also introduces a more neutral discovery and gossip layer, so users are less dependent on subscribing to a few major public relays and hoping those relays represent the whole network.
Who should software believe?
Atlas makes trust first-class data instead of leaving it entirely to client-specific heuristics.
What should rise to the top?
Atlas aims for more inspectable discovery logic than isolated relay or client decisions.
How do shared rules change?
Atlas gives communities clearer protocol and legislation layers for coordination.
Persistence and incentives are not left outside the stack
EconomicsNostr is strong at publication and relay-based distribution, but it does not by itself create a shared economy for long-term storage or relay maintenance. Operators can charge, donate, or experiment, yet that layer remains fragmented.
Atlas is designed to connect useful participation, availability, and governance more explicitly, so important data has stronger reasons to stay online.
These systems choose different scopes
Its minimal core makes it easy to implement and difficult for one company to own.
More of identity, discovery, moderation, and persistence still gets pushed outward to apps and relays.
Identity, typed records, trust, governance, and incentives sit closer to the protocol itself.
Nostr is an elegant event protocol. Atlas is trying to be a fuller application protocol.
