Vault & Proof Identifiers

Zero-knowledge proofs (ZKPs) are malleable; from two valid ZKPs, it is impossible to know whether they were generated from the same secret inputs. This means that a user can create an infinite number of valid ZKPs—posing challenges for Sybil-resistance. The question of how to limit an individual user to one valid ZKP then arises.
As a result, zero-knowledge protocols such as Tornado Cash or Semaphore use a nullifier to prevent certain effects from occurring more than once (i.e, double spends). The nullifier is typically a unique ID inherited from the secret a user provides during ZK proving schemes. ZKPs allow users to prove this connection without revealing how it is true.
Sismo uses the notion of a Vault or Proof Identifier to ensure the associated user has not already accessed certain features. These identifiers are used by applications to substantiate a user or proof’s origin in a privacy-preserving manner. This prevents users from minting duplicate Badges or accessing gated features via Sismo Connect multiple times when unintended.

Vault Identifier

Sismo revolves around personal data aggregated in a user’s Data Vault. Users can leverage data stored in their Vault via Sismo Connect, a privacy-preserving single sign-on (SSO) method. This is made possible with a Vault Identifier—an anonymous app-specific identifier that can be utilized as an in-app user ID.
A Vault Identifier is a unique deterministic number that can only be computed by the associated Vault. The uniqueness of a given Vault Identifier is ensured by its generation method. It is generated by hashing (Poseidon) the following public inputs in a zk-SNARK circuit run entirely in the user’s browser:
  • The VaultSecret generated when adding a Data Source to the Data Vault
  • The appId of the application requesting proof
Due to the one-way nature of hashing, it’s impossible to reverse-engineer a Vault Identifier and link it to the associated Vault. Furthermore, the entire set of Vault Identifiers is downloaded in a user’s browser to prevent potential correlation. A given Vault Identifier can only be generated by the owner of the associated Vault.

Proof Identifiers

Alternatively, developers can opt to accept Proof Identifiers from their users instead of Vault Identifiers. A Proof Identifier limits services to a single proof instead of a single Vault—allowing applications to have multiple services without tracking their users. It is generated from a combination of public and secret data.
The requestIdentifier corresponds to a user’s request to prove group membership to a specific application. It is obtained by hashing (keccak256) the following public inputs:
  • The appId of the application requesting proof
  • The Namespace of the application’s specific service (optional)
  • The groupId of the group the user is proving membership in
  • The groupTimestamp of the group the user is proving membership in
The groupId and groupTimestamp form the group snapshot—the unique identifier of the group the user is proving membership in.
A Proof Identifier is subsequently created by hashing (Poseidon) the following inputs inside a zk-SNARK circuit run entirely in the user’s browser:
  • The previously generated requestIdentifier
  • A user’s secret associated with the sourceAccount (private key or commitment)
The Commitment Mapper is an off-chain service that privately associates a unique commitment with a source account—ensuring users have a unique proofIdentifier derived from their address.

Verified ZKPs

When participating in proving schemes, users generate ZKPs to prove ownership of data that categorizes them into specific groups. Applications can use Sismo Connect to verify ZKPs generated to prove ownership of accounts in a user’s Data Vault.
ZKPs are verified after users provide them to Sismo’s verifier, which authenticates user proofs in a decentralized manner. The verifier can be integrated into on-chain smart contracts or off-chain applications. After verifying ZKPs provided by users, the verifier outputs a Vault or Proof Identifier, depending on the application's preferences. These identifiers can be used to control access to the application in question and are stored either in on-chain smart contracts or in off-chain databases.

Leveraging VaultId & Proof Identifiers

After storing the relevant identifier, applications can decide how to use it. VaultId and Proof Identifiers can be leveraged by applications in the following ways:
  • Nullify source accounts
  • Restrict source accounts
Nullifying a source account utilizes the identifier to ensure source accounts can not be used more than once, e.g., to mint duplicate Badges. Restricting a source account puts a limit on the number of times a source account can access a gated feature, e.g., users could claim a ticket for an event twice to bring an extra person.
As applications store the identifier, they can choose how many times a specific ZKP can be used. If a Vault or Proof Identifier exists in an application’s database (on-chain or off-chain), users can not access specific features more times than intended.