Struct Validator 

pub struct Validator { /* private fields */ }

The main validator that runs evaluation loops.

Orchestrates the full validation pipeline: loading the test corpus, selecting queries, evaluating contributors, updating ratings. Peer testing requires a TorRuntime to be passed in.

Implementations

impl Validator

pub fn new( config: ValidatorConfig, arweave: Arc<ArweaveClient>, signing_key: SigningKey, model_hash: [u8; 32], ) -> Self

Create a new validator.

pub fn contributor_ratings( &self, ) -> &Arc<RwLock<HashMap<[u8; 32], ContributorRating>>>

Get a read reference to the contributor ratings.

pub fn peer_ratings(&self) -> &Arc<RwLock<HashMap<[u8; 16], PeerRating>>>

Get a read reference to the peer ratings.

pub fn node_id(&self) -> [u8; 16]

Get this validator’s node ID.

pub async fn load_corpus(&self) -> Result<(), ValidatorError>

Load or refresh the test corpus.

Uses the local file if configured, otherwise downloads from Arweave.

pub async fn run_evaluation_cycle( &self, ) -> Result<Vec<EvaluationResult>, ValidatorError>

Run a single evaluation cycle for the current validation window.

This is the core validation loop:

1. Ensure corpus is loaded
2. Select test queries for the current window
3. Discover contributors on Arweave
4. Evaluate each contributor
5. Update contributor ratings

Returns the evaluation results for optional publication to Arweave.

pub async fn test_live_peers( &self, tor_runtime: &TorRuntime, ) -> Result<Vec<PeerTestResult>, ValidatorError>

Test live peers by connecting over Tor and running test queries.

Requires a running TorRuntime. Returns the peer test results for optional publication to Arweave.

pub async fn run_loop( &self, tor_runtime: Option<&TorRuntime>, ) -> Result<(), ValidatorError>

Run the validator loop continuously.

Runs evaluation cycles at the configured interval, sleeping between cycles. If peer testing is enabled and a TorRuntime is provided, also tests live peers each cycle.