Struct EvaluationService 

pub struct EvaluationService { /* private fields */ }

Implementations

impl EvaluationService

pub fn new(repositories: EvalRepositories, ai: DynAiProvider) -> Self

ai must be the auditing provider (the AiService implementation): judge isolation and cost accounting depend on generate persisting every request to ai_requests with a job actor.

pub async fn run_judge( &self, request: RunRequest, ) -> Result<(EvalRunId, LoopReport)>

One full auto-improve pass; returns the run id and its report.

pub async fn replay_failures( &self, source_run: &EvalRunId, request: RunRequest, ) -> Result<(EvalRunId, LoopReport)>

Re-runs a previous run’s failing requests as a new replay-kind run.

pub async fn list_runs(&self, limit: i64) -> Result<Vec<EvalRun>>

pub async fn promote_case(&self, params: &NewCaseParams) -> Result<EvalCaseId>

pub async fn get_run(&self, run_id: &EvalRunId) -> Result<EvalRun>

pub async fn list_results(&self, run_id: &EvalRunId) -> Result<Vec<EvalResult>>

pub async fn sample(&self, filter: &SampleFilter) -> Result<Vec<SampledRequest>>

Trait Implementations

impl Clone for EvaluationService

fn clone(&self) -> EvaluationService

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for EvaluationService

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.