🪢 Open source agent evals & observability: Trace, evaluate, and improve LLM applications with one open platform.
TypeScript analyticsautogenevaluationlangchainlarge-language-modelsllama-indexllmllm-evaluationllm-observabilityllmopsmonitoringobservabilityopen-sourceopenaiplaygroundprompt-engineeringprompt-managementself-hostedycombinator
import { useMemo } from "react";
import { useHasProjectAccess } from "@/src/features/rbac/utils/checkProjectAccess";
import { api } from "@/src/utils/api";
export type ExperimentNameOption = {
experimentId: string;
experimentName: string;
datasetId: string | null;
datasetName: string | null;
startTime: Date;
};
type UseExperimentNamesResponse = {
experimentNames: ExperimentNameOption[];
isLoading: boolean;
};
export function useExperimentNames({
projectId,
}: {
projectId: string;
}): UseExperimentNamesResponse {
const hasExperimentsReadAccess = useHasProjectAccess({
projectId,
scope: "promptExperiments:read",
});
const { data, isLoading } = api.experiments.byProjectId.useQuery(
{
projectId,
},
// `projectId` arrives with `router.query` after hydration; without it in
// the guard the query can fire with `undefined` and zod rejects it.
{ enabled: Boolean(projectId) && hasExperimentsReadAccess },
);
// Newest first: the run worth comparing against is a recent one, not the one
// the alphabet happens to put on top. Copied before sorting so the query
// cache is not mutated in place.
const experimentNames = useMemo(
() =>
[...(data?.experimentNames ?? [])].sort(
(a, b) => b.startTime.getTime() - a.startTime.getTime(),
),
[data?.experimentNames],
);
return { experimentNames, isLoading };
}