In the context of Intentional Theory, an identifiable agent is defined as any process—human, organizational, or automated—that possesses a private key and can use it to sign events in a verifiable manner.
This cryptographic capability establishes accountability and authorship within the system, enabling the agent to make intentional declarations that can be traced to a consistent identity (represented by the corresponding public key).
The agent’s identity is not determined by external attributes such as name or origin, but by its ability to produce signatures that can be publicly verified. This definition ensures that intention is always anchored to a persistent, accountable source, regardless of whether the agent is a person, a group, or a machine.
The Nostr protocol uses public keys based on BIP340, the specification for Schnorr signatures over the secp256k1 elliptic curve. This cryptographic scheme enables what can be described as algorithmic accountability—a form of verifiable authorship grounded in mathematics rather than trust in institutions or intermediaries.
Each event in Nostr is signed with a private key, and the corresponding public key acts as a persistent, pseudonymous identifier. Because Schnorr signatures are deterministic and non-malleable, anyone can independently verify that a given event was authored by the holder of a specific private key, without revealing that key or requiring third-party validation.
This form of accountability is algorithmic because it does not rely on identity verification in the traditional sense (e.g., real names, biometrics, or centralized certificates), but instead on cryptographic proof. The system ensures that only the agent in possession of the private key could have authored the event, providing a robust foundation for intentional communication. It also ensures that identities cannot be spoofed or forged, as signatures that do not match the public key are invalid.
In this way, Nostr provides a substrate where intention and authorship are inseparable, persistent, and computationally verifiable—enabling decentralized trust without reliance on external authorities.