Skip to main content

remem/eval/
coding_bench.rs

1mod artifact;
2mod audit_contract;
3mod capture_projection;
4mod condition;
5mod curator;
6mod dry_run;
7mod e2e;
8mod failure;
9mod fixture;
10mod isolation;
11mod live_approval;
12mod preflight;
13mod process;
14mod run_plan;
15mod runner;
16mod score;
17#[cfg(test)]
18mod tests;
19mod types;
20
21pub use artifact::{
22    build_remem_contract_snapshot, validate_contract_snapshots, CodingBenchCondition,
23    CodingBenchConditionReport, CodingBenchMemoryContractStatus, CodingBenchReport,
24    CodingBenchRunMetrics, CodingBenchRunReport, CodingBenchRunScoreEvidence,
25    CodingBenchScoreCommandEvidence, RememContractHealth, RememContractSnapshot,
26    RememContractWarning, RememInjectedMemoryAuditSnapshot, RememStalenessHandlingSnapshot,
27    RememTemporalFactEligibilitySnapshot, RememUsageFeedbackCoverageSnapshot,
28    CODING_AGENT_AB_SPEC_PATH, CURRENT_MEMORY_CONTRACT_SPEC_PATH, MIN_RUNS_PER_CONDITION,
29};
30pub(crate) use audit_contract::verify_snapshot_against_persisted_injection;
31pub use audit_contract::{
32    verify_context_audit_snapshot, RememContextAuditSnapshot, RememContextAuditStatus,
33};
34pub use runner::{dry_run_plan, run_coding_bench};
35pub use types::CodingBenchOptions;
36
37pub(crate) fn verify_live_approval_json(options: &CodingBenchOptions) -> anyhow::Result<String> {
38    live_approval::verify_live_approval_json(options)
39}