@iblai/iblai-js 1.26.2 → 1.26.4

package/dist/web-containers/source/index.esm.js

@@ -62710,7 +62710,9 @@ var profileAccount$3 = {
 	descAdvanced: "Configure advanced organization settings.",
 	descBasic: "Manage your basic account information and preferences.",
 	descSocial: "Connect and manage your social media accounts.",
-	descSecurity: "Update your security settings and password."
+	descSecurity: "Update your security settings and password.",
+	tabBenchmarks: "Benchmarks",
+	descBenchmarks: "Build benchmark suites of test inputs to evaluate agents."
 };
 var profileAddCreditsModal$3 = {
 	title: "Add Credits",
@@ -64712,6 +64714,220 @@ var ltiTabLabels$3 = {
 		toolError: "Something went wrong with the LTI tool"
 	}
 };
+var evaluationTabLabels$3 = {
+	header: {
+		title: "Evals",
+		description: "Run this agent against a benchmark and score the responses.",
+		infoBox: "Measure how well your agent performs. Run it against a benchmark — a set of test questions — and every response gets scored, so you can catch weak spots and track improvements as you refine the agent."
+	},
+	benchmarkPicker: {
+		placeholder: "Select a benchmark",
+		searchPlaceholder: "Search benchmarks...",
+		empty: "No benchmarks found"
+	},
+	actions: {
+		newExperiment: "New Evaluation",
+		manageBenchmarks: "Manage benchmarks",
+		viewResults: "View results",
+		evaluate: "New review",
+		exportCsv: "Export CSV",
+		exporting: "Exporting...",
+		checkStatus: "Check status",
+		checkingStatus: "Checking...",
+		notReadyHint: "Available once the evaluation completes",
+		"delete": "Delete",
+		rowMenuAriaLabel: "Actions for {name}"
+	},
+	checkStatusToast: {
+		pending: "Evaluation is still running",
+		completed: "Evaluation completed",
+		failed: "Evaluation failed"
+	},
+	table: {
+		experimentColumn: "EVALUATION",
+		statusColumn: "STATUS",
+		initiatedByColumn: "INITIATED BY",
+		createdColumn: "CREATED",
+		actionsColumn: "ACTIONS",
+		emptyState: "No evaluations yet for this agent on benchmark \"{dataset}\"."
+	},
+	emptyBenchmarks: "No benchmarks yet. Use \"Manage benchmarks\" above to create one.",
+	exportErrorToast: "Failed to export evaluation",
+	startExperimentModal: {
+		title: "New Evaluation",
+		description: "Run this agent against every item in the selected benchmark. Results show up here once the run completes in the background.",
+		benchmarkLabel: "Benchmark",
+		experimentNameLabel: "Evaluation name",
+		experimentNamePlaceholder: "Auto-generated if left blank",
+		cancel: "Cancel",
+		start: "Start Evaluation",
+		successToast: "Evaluation started",
+		errorToast: "Failed to start evaluation"
+	},
+	manageBenchmarksModal: {
+		title: "Manage benchmarks",
+		description: "Create and edit tenant-wide benchmarks without leaving the agent settings."
+	},
+	detailModal: {
+		benchmarkLabel: "Benchmark",
+		startedPrefix: "started",
+		tracesSuffixSingular: "trace produced",
+		tracesSuffixPlural: "traces produced",
+		waitingHint: "Waiting for the evaluation to produce traces. This page auto-refreshes.",
+		noScoresHint: "No scores yet",
+		newReview: "New review",
+		scoreNameLabel: "Score name",
+		scoreNamePlaceholder: "accuracy",
+		scoreValueLabel: "Value",
+		scoreValuePlaceholder: "0.0",
+		scoreCommentLabel: "Comment",
+		scoreCommentPlaceholder: "Optional",
+		addScore: "Add",
+		scoreInputError: "Score name and numeric value are required",
+		scoreRangeError: "Score value must be between 0 and 1",
+		scoreAddedToast: "Score added",
+		scoreErrorToast: "Failed to add score",
+		scoreRemovedToast: "Score removed",
+		scoreRemoveErrorToast: "Failed to remove score",
+		traceLabel: "Trace",
+		itemLabel: "item",
+		removeScoreAriaLabel: "Remove score",
+		inputLabel: "Input",
+		expectedOutputLabel: "Expected output",
+		actualOutputLabel: "Actual output",
+		errorLabel: "Error",
+		noOutputHint: "No output yet",
+		scoresHeading: "Scores",
+		addScoreHeading: "Add a score",
+		reviewsHeading: "Reviews",
+		reviewsHint: "Reviews are LLM-as-Judge runs that score every trace in this experiment. Scores from completed reviews appear under each item below.",
+		reviewsRunningLabel: "running",
+		reviewsCompletedLabel: "complete",
+		reviewsFailedLabel: "failed",
+		reviewUnnamedScore: "Untitled score",
+		reviewFinishedPrefix: "finished",
+		reviewStartedPrefix: "started",
+		reviewStatus: {
+			pending: "Queued",
+			running: "Running",
+			completed: "Complete",
+			failed: "Failed"
+		},
+		refresh: "Refresh",
+		reviewDetailsAriaLabel: "Toggle review details",
+		reviewCriteriaLabel: "Criteria",
+		reviewScoresProducedLabel: "Scores produced",
+		reviewNoScoresYet: "No scores recorded for this review yet.",
+		scoresSuffixSingular: "score",
+		scoresSuffixPlural: "scores"
+	},
+	evaluateModal: {
+		title: "Evaluate",
+		description: "Use an LLM to score every trace in this evaluation.",
+		criteriaLabel: "Criteria",
+		criteriaPlaceholder: "e.g. Evaluate the response on:\n1. Accuracy\n2. Completeness\n3. Clarity\nWeight accuracy most heavily.",
+		scoreNameLabel: "Score name",
+		scoreNamePlaceholder: "e.g. quality",
+		maxConcurrencyLabel: "Max concurrency",
+		llmLabel: "LLM",
+		selectLlmPlaceholder: "Select LLM",
+		cancel: "Cancel",
+		submit: "Evaluate",
+		validationError: "Criteria, score name and LLM are required",
+		successToast: "Evaluation started",
+		errorToast: "Failed to start evaluation",
+		pickerTitle: "Select LLM",
+		pickerDescription: "Pick a provider then a model.",
+		pickerSelectorAriaLabel: "LLM Model Selector",
+		pickerSearchPlaceholder: "Search providers...",
+		pickerEmpty: "No providers",
+		pickerLogoAlt: "{name} logo"
+	},
+	deleteModal: {
+		title: "Delete Evaluation",
+		confirmation: "Are you sure you want to delete this evaluation? This action cannot be undone.",
+		cancel: "Cancel",
+		confirm: "Delete",
+		confirming: "Deleting...",
+		successToast: "Evaluation deleted",
+		errorToast: "Failed to delete evaluation"
+	}
+};
+var benchmarksIndex$3 = {
+	searchPlaceholder: "Search benchmarks",
+	searchAriaLabel: "Search benchmarks",
+	newBenchmark: "New Benchmark",
+	nameColumn: "Name",
+	descriptionColumn: "Description",
+	createdColumn: "Created",
+	actionsColumn: "Actions",
+	viewItemsAriaLabel: "View items in {name}",
+	loading: "Loading...",
+	empty: "No benchmarks found"
+};
+var benchmarksCreateBenchmarkModal$3 = {
+	title: "New Benchmark",
+	description: "Create a benchmark to group test inputs for agent evaluation runs.",
+	nameLabel: "Name",
+	namePlaceholder: "qa-eval-v1",
+	descriptionLabel: "Description",
+	descriptionPlaceholder: "What is this benchmark for?",
+	nameRequired: "Benchmark name is required",
+	cancel: "Cancel",
+	create: "Create",
+	successToast: "Benchmark created",
+	errorToast: "Failed to create benchmark"
+};
+var benchmarksBenchmarkItemsModal$3 = {
+	description: "Question/answer pairs used to evaluate agents against this benchmark.",
+	addItems: "Add Q&A",
+	questionColumn: "Question",
+	answerColumn: "Answer",
+	actionsColumn: "Actions",
+	deleteItemAriaLabel: "Delete Q&A",
+	empty: "No Q&A in this benchmark yet"
+};
+var benchmarksDeleteItemConfirm$3 = {
+	title: "Delete Q&A",
+	confirmation: "Are you sure you want to delete this Q&A pair? This action cannot be undone.",
+	cancel: "Cancel",
+	"delete": "Delete",
+	deleting: "Deleting...",
+	successToast: "Q&A deleted",
+	errorToast: "Failed to delete Q&A"
+};
+var benchmarksAddItemsModal$3 = {
+	title: "Add Q&A to {name}",
+	description: "Add question/answer pairs directly or upload a CSV file.",
+	manualTab: "Manual",
+	csvTab: "CSV Upload",
+	questionLabel: "Question",
+	questionPlaceholder: "What is machine learning?",
+	answerLabel: "Answer",
+	answerPlaceholder: "Optional",
+	removeRowAriaLabel: "Remove row",
+	addRow: "Add row",
+	cancel: "Cancel",
+	add: "Add",
+	rowRequired: "Add at least one row with a question",
+	addedToast: "{count, plural, one {Added # Q&A pair} other {Added # Q&A pairs}}",
+	addErrorToast: "Failed to add Q&A",
+	csvFormatTitle: "CSV format",
+	csvRuleEncoding: "UTF-8 encoded",
+	csvRuleRequiredPrefix: "Required column:",
+	csvRuleRequiredSuffix: "(the question)",
+	csvRuleOptionalPrefix: "Optional column:",
+	csvRuleOptionalSuffix: "(the answer)",
+	csvRuleMaxSize: "Max file size 10 MB, up to 10,000 rows",
+	csvRuleBlankSkipped: "Rows with a blank question are skipped",
+	downloadSample: "Download sample CSV",
+	csvFileLabel: "CSV file",
+	selectedFile: "Selected: {name} ({size} KB)",
+	csvRequired: "Select a CSV file",
+	uploadedToast: "{count, plural, one {Uploaded # Q&A pair} other {Uploaded # Q&A pairs}}",
+	uploadErrorToast: "Failed to upload CSV",
+	upload: "Upload"
+};
 var platformSidebar$3 = {
 	collapseSidebar: "Collapse sidebar",
 	expandSidebar: "Expand sidebar",
@@ -64974,6 +65190,12 @@ var en = {
 	workflowsDeleteWorkflowModal: workflowsDeleteWorkflowModal$3,
 	workflowsToolDialogs: workflowsToolDialogs$3,
 	ltiTabLabels: ltiTabLabels$3,
+	evaluationTabLabels: evaluationTabLabels$3,
+	benchmarksIndex: benchmarksIndex$3,
+	benchmarksCreateBenchmarkModal: benchmarksCreateBenchmarkModal$3,
+	benchmarksBenchmarkItemsModal: benchmarksBenchmarkItemsModal$3,
+	benchmarksDeleteItemConfirm: benchmarksDeleteItemConfirm$3,
+	benchmarksAddItemsModal: benchmarksAddItemsModal$3,
 	platformSidebar: platformSidebar$3,
 	platformNavbar: platformNavbar$3
 };
@@ -68092,7 +68314,9 @@ var profileAccount$2 = {
 	descAdvanced: "Configurer les paramètres avancés de l'organisation.",
 	descBasic: "Gérer vos informations de compte de base et vos préférences.",
 	descSocial: "Connecter et gérer vos comptes de réseaux sociaux.",
-	descSecurity: "Mettre à jour vos paramètres de sécurité et votre mot de passe."
+	descSecurity: "Mettre à jour vos paramètres de sécurité et votre mot de passe.",
+	tabBenchmarks: "Benchmarks",
+	descBenchmarks: "Créez des suites de benchmarks d'entrées de test pour évaluer les agents."
 };
 var profileAddCreditsModal$2 = {
 	title: "Ajouter des crédits",
@@ -70094,6 +70318,220 @@ var ltiTabLabels$2 = {
 		toolError: "Un problème est survenu avec l'outil LTI"
 	}
 };
+var evaluationTabLabels$2 = {
+	header: {
+		title: "Évaluations",
+		description: "Exécutez cet agent sur un benchmark et notez les réponses.",
+		infoBox: "Mesurez les performances réelles de votre agent. Exécutez-le sur un benchmark — un ensemble de questions de test — et chaque réponse est notée, pour repérer les points faibles et suivre les progrès à mesure que vous perfectionnez l'agent."
+	},
+	benchmarkPicker: {
+		placeholder: "Sélectionner un benchmark",
+		searchPlaceholder: "Rechercher des benchmarks...",
+		empty: "Aucun benchmark trouvé"
+	},
+	actions: {
+		newExperiment: "Nouvelle évaluation",
+		manageBenchmarks: "Gérer les benchmarks",
+		viewResults: "Voir les résultats",
+		evaluate: "Nouvelle revue",
+		exportCsv: "Exporter en CSV",
+		exporting: "Exportation...",
+		checkStatus: "Vérifier le statut",
+		checkingStatus: "Vérification...",
+		notReadyHint: "Disponible une fois l'évaluation terminée",
+		"delete": "Supprimer",
+		rowMenuAriaLabel: "Actions pour {name}"
+	},
+	checkStatusToast: {
+		pending: "L'évaluation est toujours en cours",
+		completed: "Évaluation terminée",
+		failed: "L'évaluation a échoué"
+	},
+	table: {
+		experimentColumn: "ÉVALUATION",
+		statusColumn: "STATUT",
+		initiatedByColumn: "INITIÉE PAR",
+		createdColumn: "CRÉÉE",
+		actionsColumn: "ACTIONS",
+		emptyState: "Aucune évaluation pour cet agent sur le benchmark \"{dataset}\" pour le moment."
+	},
+	emptyBenchmarks: "Aucun benchmark pour le moment. Utilisez \"Gérer les benchmarks\" ci-dessus pour en créer un.",
+	exportErrorToast: "Échec de l'exportation de l'évaluation",
+	startExperimentModal: {
+		title: "Nouvelle évaluation",
+		description: "Exécutez cet agent sur chaque élément du benchmark sélectionné. Les résultats apparaîtront ici une fois l'exécution terminée en arrière-plan.",
+		benchmarkLabel: "Benchmark",
+		experimentNameLabel: "Nom de l'évaluation",
+		experimentNamePlaceholder: "Généré automatiquement si laissé vide",
+		cancel: "Annuler",
+		start: "Lancer l'évaluation",
+		successToast: "Évaluation lancée",
+		errorToast: "Échec du lancement de l'évaluation"
+	},
+	manageBenchmarksModal: {
+		title: "Gérer les benchmarks",
+		description: "Créez et modifiez les benchmarks de l'organisation sans quitter les paramètres de l'agent."
+	},
+	detailModal: {
+		benchmarkLabel: "Benchmark",
+		startedPrefix: "démarrée",
+		tracesSuffixSingular: "trace produite",
+		tracesSuffixPlural: "traces produites",
+		waitingHint: "En attente que l'évaluation produise des traces. Cette page s'actualise automatiquement.",
+		noScoresHint: "Pas encore de scores",
+		newReview: "Nouvelle revue",
+		scoreNameLabel: "Nom du score",
+		scoreNamePlaceholder: "précision",
+		scoreValueLabel: "Valeur",
+		scoreValuePlaceholder: "0.0",
+		scoreCommentLabel: "Commentaire",
+		scoreCommentPlaceholder: "Facultatif",
+		addScore: "Ajouter",
+		scoreInputError: "Le nom du score et une valeur numérique sont requis",
+		scoreRangeError: "La valeur du score doit être comprise entre 0 et 1",
+		scoreAddedToast: "Score ajouté",
+		scoreErrorToast: "Échec de l'ajout du score",
+		scoreRemovedToast: "Score supprimé",
+		scoreRemoveErrorToast: "Échec de la suppression du score",
+		traceLabel: "Trace",
+		itemLabel: "élément",
+		removeScoreAriaLabel: "Supprimer le score",
+		inputLabel: "Entrée",
+		expectedOutputLabel: "Sortie attendue",
+		actualOutputLabel: "Sortie réelle",
+		errorLabel: "Erreur",
+		noOutputHint: "Pas encore de sortie",
+		scoresHeading: "Scores",
+		addScoreHeading: "Ajouter un score",
+		reviewsHeading: "Revues",
+		reviewsHint: "Les revues sont des exécutions LLM-juge qui notent chaque trace de cette expérience. Les scores des revues terminées apparaissent sous chaque élément ci-dessous.",
+		reviewsRunningLabel: "en cours",
+		reviewsCompletedLabel: "terminées",
+		reviewsFailedLabel: "échouées",
+		reviewUnnamedScore: "Score sans titre",
+		reviewFinishedPrefix: "terminée",
+		reviewStartedPrefix: "démarrée",
+		reviewStatus: {
+			pending: "En file d'attente",
+			running: "En cours",
+			completed: "Terminée",
+			failed: "Échouée"
+		},
+		refresh: "Actualiser",
+		reviewDetailsAriaLabel: "Afficher ou masquer les détails de la revue",
+		reviewCriteriaLabel: "Critères",
+		reviewScoresProducedLabel: "Scores produits",
+		reviewNoScoresYet: "Aucun score enregistré pour cette revue pour le moment.",
+		scoresSuffixSingular: "score",
+		scoresSuffixPlural: "scores"
+	},
+	evaluateModal: {
+		title: "Évaluer",
+		description: "Utilisez un LLM pour noter chaque trace de cette évaluation.",
+		criteriaLabel: "Critères",
+		criteriaPlaceholder: "ex. Évaluez la réponse selon :\n1. Exactitude\n2. Exhaustivité\n3. Clarté\nAccordez le plus de poids à l'exactitude.",
+		scoreNameLabel: "Nom du score",
+		scoreNamePlaceholder: "ex. qualité",
+		maxConcurrencyLabel: "Concurrence maximale",
+		llmLabel: "LLM",
+		selectLlmPlaceholder: "Sélectionner un LLM",
+		cancel: "Annuler",
+		submit: "Évaluer",
+		validationError: "Les critères, le nom du score et le LLM sont requis",
+		successToast: "Évaluation lancée",
+		errorToast: "Échec du lancement de l'évaluation",
+		pickerTitle: "Sélectionner un LLM",
+		pickerDescription: "Choisissez un fournisseur puis un modèle.",
+		pickerSelectorAriaLabel: "Sélecteur de modèle LLM",
+		pickerSearchPlaceholder: "Rechercher des fournisseurs...",
+		pickerEmpty: "Aucun fournisseur",
+		pickerLogoAlt: "Logo de {name}"
+	},
+	deleteModal: {
+		title: "Supprimer l'évaluation",
+		confirmation: "Voulez-vous vraiment supprimer cette évaluation ? Cette action est irréversible.",
+		cancel: "Annuler",
+		confirm: "Supprimer",
+		confirming: "Suppression...",
+		successToast: "Évaluation supprimée",
+		errorToast: "Échec de la suppression de l'évaluation"
+	}
+};
+var benchmarksIndex$2 = {
+	searchPlaceholder: "Rechercher des benchmarks",
+	searchAriaLabel: "Rechercher des benchmarks",
+	newBenchmark: "Nouveau benchmark",
+	nameColumn: "Nom",
+	descriptionColumn: "Description",
+	createdColumn: "Créé",
+	actionsColumn: "Actions",
+	viewItemsAriaLabel: "Voir les éléments de {name}",
+	loading: "Chargement...",
+	empty: "Aucun benchmark trouvé"
+};
+var benchmarksCreateBenchmarkModal$2 = {
+	title: "Nouveau benchmark",
+	description: "Créez un benchmark pour regrouper des entrées de test pour les évaluations de l'agent.",
+	nameLabel: "Nom",
+	namePlaceholder: "qa-eval-v1",
+	descriptionLabel: "Description",
+	descriptionPlaceholder: "À quoi sert ce benchmark ?",
+	nameRequired: "Le nom du benchmark est requis",
+	cancel: "Annuler",
+	create: "Créer",
+	successToast: "Benchmark créé",
+	errorToast: "Échec de la création du benchmark"
+};
+var benchmarksBenchmarkItemsModal$2 = {
+	description: "Paires question-réponse utilisées pour évaluer les agents sur ce benchmark.",
+	addItems: "Ajouter des questions-réponses",
+	questionColumn: "Question",
+	answerColumn: "Réponse",
+	actionsColumn: "Actions",
+	deleteItemAriaLabel: "Supprimer la question-réponse",
+	empty: "Aucune question-réponse dans ce benchmark pour le moment"
+};
+var benchmarksDeleteItemConfirm$2 = {
+	title: "Supprimer la question-réponse",
+	confirmation: "Voulez-vous vraiment supprimer cette paire question-réponse ? Cette action est irréversible.",
+	cancel: "Annuler",
+	"delete": "Supprimer",
+	deleting: "Suppression...",
+	successToast: "Question-réponse supprimée",
+	errorToast: "Échec de la suppression de la question-réponse"
+};
+var benchmarksAddItemsModal$2 = {
+	title: "Ajouter des questions-réponses à {name}",
+	description: "Ajoutez des paires question-réponse directement ou importez un fichier CSV.",
+	manualTab: "Manuel",
+	csvTab: "Import CSV",
+	questionLabel: "Question",
+	questionPlaceholder: "Qu'est-ce que l'apprentissage automatique ?",
+	answerLabel: "Réponse",
+	answerPlaceholder: "Facultatif",
+	removeRowAriaLabel: "Supprimer la ligne",
+	addRow: "Ajouter une ligne",
+	cancel: "Annuler",
+	add: "Ajouter",
+	rowRequired: "Ajoutez au moins une ligne avec une question",
+	addedToast: "{count, plural, one {# paire question-réponse ajoutée} other {# paires question-réponse ajoutées}}",
+	addErrorToast: "Échec de l'ajout des questions-réponses",
+	csvFormatTitle: "Format CSV",
+	csvRuleEncoding: "Encodé en UTF-8",
+	csvRuleRequiredPrefix: "Colonne requise :",
+	csvRuleRequiredSuffix: "(la question)",
+	csvRuleOptionalPrefix: "Colonne facultative :",
+	csvRuleOptionalSuffix: "(la réponse)",
+	csvRuleMaxSize: "Taille maximale du fichier 10 Mo, jusqu'à 10 000 lignes",
+	csvRuleBlankSkipped: "Les lignes sans question sont ignorées",
+	downloadSample: "Télécharger un CSV d’exemple",
+	csvFileLabel: "Fichier CSV",
+	selectedFile: "Sélectionné : {name} ({size} Ko)",
+	csvRequired: "Sélectionnez un fichier CSV",
+	uploadedToast: "{count, plural, one {# paire question-réponse importée} other {# paires question-réponse importées}}",
+	uploadErrorToast: "Échec de l'import du CSV",
+	upload: "Importer"
+};
 var platformSidebar$2 = {
 	collapseSidebar: "Réduire la barre latérale",
 	expandSidebar: "Développer la barre latérale",
@@ -70356,6 +70794,12 @@ var fr = {
 	workflowsDeleteWorkflowModal: workflowsDeleteWorkflowModal$2,
 	workflowsToolDialogs: workflowsToolDialogs$2,
 	ltiTabLabels: ltiTabLabels$2,
+	evaluationTabLabels: evaluationTabLabels$2,
+	benchmarksIndex: benchmarksIndex$2,
+	benchmarksCreateBenchmarkModal: benchmarksCreateBenchmarkModal$2,
+	benchmarksBenchmarkItemsModal: benchmarksBenchmarkItemsModal$2,
+	benchmarksDeleteItemConfirm: benchmarksDeleteItemConfirm$2,
+	benchmarksAddItemsModal: benchmarksAddItemsModal$2,
 	platformSidebar: platformSidebar$2,
 	platformNavbar: platformNavbar$2
 };
@@ -73474,7 +73918,9 @@ var profileAccount$1 = {
 	descAdvanced: "Configurar los ajustes avanzados de la organización.",
 	descBasic: "Gestionar su información básica de cuenta y preferencias.",
 	descSocial: "Conectar y gestionar sus cuentas de redes sociales.",
-	descSecurity: "Actualizar su configuración de seguridad y contraseña."
+	descSecurity: "Actualizar su configuración de seguridad y contraseña.",
+	tabBenchmarks: "Benchmarks",
+	descBenchmarks: "Cree suites de benchmarks con entradas de prueba para evaluar agentes."
 };
 var profileAddCreditsModal$1 = {
 	title: "Agregar créditos",
@@ -75476,6 +75922,220 @@ var ltiTabLabels$1 = {
 		toolError: "Algo salió mal con la herramienta LTI"
 	}
 };
+var evaluationTabLabels$1 = {
+	header: {
+		title: "Evaluaciones",
+		description: "Ejecuta este agente contra un benchmark y puntúa las respuestas.",
+		infoBox: "Mide el rendimiento real de tu agente. Ejecútalo contra un benchmark — un conjunto de preguntas de prueba — y cada respuesta se puntúa, para detectar puntos débiles y seguir las mejoras a medida que ajustas el agente."
+	},
+	benchmarkPicker: {
+		placeholder: "Selecciona un benchmark",
+		searchPlaceholder: "Buscar benchmarks...",
+		empty: "No se encontraron benchmarks"
+	},
+	actions: {
+		newExperiment: "Nueva evaluación",
+		manageBenchmarks: "Gestionar benchmarks",
+		viewResults: "Ver resultados",
+		evaluate: "Nueva revisión",
+		exportCsv: "Exportar CSV",
+		exporting: "Exportando...",
+		checkStatus: "Comprobar estado",
+		checkingStatus: "Comprobando...",
+		notReadyHint: "Disponible cuando la evaluación termine",
+		"delete": "Eliminar",
+		rowMenuAriaLabel: "Acciones para {name}"
+	},
+	checkStatusToast: {
+		pending: "La evaluación sigue en curso",
+		completed: "Evaluación completada",
+		failed: "La evaluación falló"
+	},
+	table: {
+		experimentColumn: "EVALUACIÓN",
+		statusColumn: "ESTADO",
+		initiatedByColumn: "INICIADA POR",
+		createdColumn: "CREADA",
+		actionsColumn: "ACCIONES",
+		emptyState: "Aún no hay evaluaciones de este agente en el benchmark \"{dataset}\"."
+	},
+	emptyBenchmarks: "Aún no hay benchmarks. Usa \"Gestionar benchmarks\" arriba para crear uno.",
+	exportErrorToast: "No se pudo exportar la evaluación",
+	startExperimentModal: {
+		title: "Nueva evaluación",
+		description: "Ejecuta este agente contra cada elemento del benchmark seleccionado. Los resultados aparecerán aquí cuando la ejecución termine en segundo plano.",
+		benchmarkLabel: "Benchmark",
+		experimentNameLabel: "Nombre de la evaluación",
+		experimentNamePlaceholder: "Se genera automáticamente si se deja en blanco",
+		cancel: "Cancelar",
+		start: "Iniciar evaluación",
+		successToast: "Evaluación iniciada",
+		errorToast: "No se pudo iniciar la evaluación"
+	},
+	manageBenchmarksModal: {
+		title: "Gestionar benchmarks",
+		description: "Crea y edita los benchmarks de toda la organización sin salir de la configuración del agente."
+	},
+	detailModal: {
+		benchmarkLabel: "Benchmark",
+		startedPrefix: "iniciada",
+		tracesSuffixSingular: "traza generada",
+		tracesSuffixPlural: "trazas generadas",
+		waitingHint: "Esperando a que la evaluación genere trazas. Esta página se actualiza automáticamente.",
+		noScoresHint: "Aún no hay puntuaciones",
+		newReview: "Nueva revisión",
+		scoreNameLabel: "Nombre de la puntuación",
+		scoreNamePlaceholder: "precisión",
+		scoreValueLabel: "Valor",
+		scoreValuePlaceholder: "0.0",
+		scoreCommentLabel: "Comentario",
+		scoreCommentPlaceholder: "Opcional",
+		addScore: "Añadir",
+		scoreInputError: "Se requieren el nombre de la puntuación y un valor numérico",
+		scoreRangeError: "El valor de la puntuación debe estar entre 0 y 1",
+		scoreAddedToast: "Puntuación añadida",
+		scoreErrorToast: "No se pudo añadir la puntuación",
+		scoreRemovedToast: "Puntuación eliminada",
+		scoreRemoveErrorToast: "No se pudo eliminar la puntuación",
+		traceLabel: "Traza",
+		itemLabel: "elemento",
+		removeScoreAriaLabel: "Eliminar puntuación",
+		inputLabel: "Entrada",
+		expectedOutputLabel: "Salida esperada",
+		actualOutputLabel: "Salida real",
+		errorLabel: "Error",
+		noOutputHint: "Aún sin salida",
+		scoresHeading: "Puntuaciones",
+		addScoreHeading: "Añadir una puntuación",
+		reviewsHeading: "Revisiones",
+		reviewsHint: "Las revisiones son ejecuciones de LLM como juez que puntúan cada traza de este experimento. Las puntuaciones de las revisiones completadas aparecen bajo cada elemento más abajo.",
+		reviewsRunningLabel: "en curso",
+		reviewsCompletedLabel: "completadas",
+		reviewsFailedLabel: "fallidas",
+		reviewUnnamedScore: "Puntuación sin título",
+		reviewFinishedPrefix: "finalizada",
+		reviewStartedPrefix: "iniciada",
+		reviewStatus: {
+			pending: "En cola",
+			running: "En curso",
+			completed: "Completada",
+			failed: "Fallida"
+		},
+		refresh: "Actualizar",
+		reviewDetailsAriaLabel: "Mostrar u ocultar detalles de la revisión",
+		reviewCriteriaLabel: "Criterios",
+		reviewScoresProducedLabel: "Puntuaciones generadas",
+		reviewNoScoresYet: "Aún no hay puntuaciones registradas para esta revisión.",
+		scoresSuffixSingular: "puntuación",
+		scoresSuffixPlural: "puntuaciones"
+	},
+	evaluateModal: {
+		title: "Evaluar",
+		description: "Usa un LLM para puntuar cada traza de esta evaluación.",
+		criteriaLabel: "Criterios",
+		criteriaPlaceholder: "p. ej. Evalúa la respuesta según:\n1. Precisión\n2. Exhaustividad\n3. Claridad\nDa más peso a la precisión.",
+		scoreNameLabel: "Nombre de la puntuación",
+		scoreNamePlaceholder: "p. ej. calidad",
+		maxConcurrencyLabel: "Concurrencia máxima",
+		llmLabel: "LLM",
+		selectLlmPlaceholder: "Seleccionar LLM",
+		cancel: "Cancelar",
+		submit: "Evaluar",
+		validationError: "Se requieren los criterios, el nombre de la puntuación y el LLM",
+		successToast: "Evaluación iniciada",
+		errorToast: "No se pudo iniciar la evaluación",
+		pickerTitle: "Seleccionar LLM",
+		pickerDescription: "Elige un proveedor y luego un modelo.",
+		pickerSelectorAriaLabel: "Selector de modelo LLM",
+		pickerSearchPlaceholder: "Buscar proveedores...",
+		pickerEmpty: "Sin proveedores",
+		pickerLogoAlt: "Logotipo de {name}"
+	},
+	deleteModal: {
+		title: "Eliminar evaluación",
+		confirmation: "¿Seguro que quieres eliminar esta evaluación? Esta acción no se puede deshacer.",
+		cancel: "Cancelar",
+		confirm: "Eliminar",
+		confirming: "Eliminando...",
+		successToast: "Evaluación eliminada",
+		errorToast: "No se pudo eliminar la evaluación"
+	}
+};
+var benchmarksIndex$1 = {
+	searchPlaceholder: "Buscar benchmarks",
+	searchAriaLabel: "Buscar benchmarks",
+	newBenchmark: "Nuevo benchmark",
+	nameColumn: "Nombre",
+	descriptionColumn: "Descripción",
+	createdColumn: "Creado",
+	actionsColumn: "Acciones",
+	viewItemsAriaLabel: "Ver elementos de {name}",
+	loading: "Cargando...",
+	empty: "No se encontraron benchmarks"
+};
+var benchmarksCreateBenchmarkModal$1 = {
+	title: "Nuevo benchmark",
+	description: "Crea un benchmark para agrupar entradas de prueba para las evaluaciones del agente.",
+	nameLabel: "Nombre",
+	namePlaceholder: "qa-eval-v1",
+	descriptionLabel: "Descripción",
+	descriptionPlaceholder: "¿Para qué es este benchmark?",
+	nameRequired: "El nombre del benchmark es obligatorio",
+	cancel: "Cancelar",
+	create: "Crear",
+	successToast: "Benchmark creado",
+	errorToast: "No se pudo crear el benchmark"
+};
+var benchmarksBenchmarkItemsModal$1 = {
+	description: "Pares de pregunta y respuesta usados para evaluar agentes con este benchmark.",
+	addItems: "Añadir preguntas y respuestas",
+	questionColumn: "Pregunta",
+	answerColumn: "Respuesta",
+	actionsColumn: "Acciones",
+	deleteItemAriaLabel: "Eliminar pregunta y respuesta",
+	empty: "Este benchmark aún no tiene preguntas y respuestas"
+};
+var benchmarksDeleteItemConfirm$1 = {
+	title: "Eliminar pregunta y respuesta",
+	confirmation: "¿Seguro que quieres eliminar este par de pregunta y respuesta? Esta acción no se puede deshacer.",
+	cancel: "Cancelar",
+	"delete": "Eliminar",
+	deleting: "Eliminando...",
+	successToast: "Pregunta y respuesta eliminadas",
+	errorToast: "No se pudo eliminar la pregunta y respuesta"
+};
+var benchmarksAddItemsModal$1 = {
+	title: "Añadir preguntas y respuestas a {name}",
+	description: "Añade pares de pregunta y respuesta directamente o sube un archivo CSV.",
+	manualTab: "Manual",
+	csvTab: "Subir CSV",
+	questionLabel: "Pregunta",
+	questionPlaceholder: "¿Qué es el aprendizaje automático?",
+	answerLabel: "Respuesta",
+	answerPlaceholder: "Opcional",
+	removeRowAriaLabel: "Eliminar fila",
+	addRow: "Añadir fila",
+	cancel: "Cancelar",
+	add: "Añadir",
+	rowRequired: "Añade al menos una fila con una pregunta",
+	addedToast: "{count, plural, one {Se añadió # par de pregunta y respuesta} other {Se añadieron # pares de pregunta y respuesta}}",
+	addErrorToast: "No se pudieron añadir las preguntas y respuestas",
+	csvFormatTitle: "Formato CSV",
+	csvRuleEncoding: "Codificado en UTF-8",
+	csvRuleRequiredPrefix: "Columna obligatoria:",
+	csvRuleRequiredSuffix: "(la pregunta)",
+	csvRuleOptionalPrefix: "Columna opcional:",
+	csvRuleOptionalSuffix: "(la respuesta)",
+	csvRuleMaxSize: "Tamaño máximo de archivo 10 MB, hasta 10,000 filas",
+	csvRuleBlankSkipped: "Las filas sin pregunta se omiten",
+	downloadSample: "Descargar CSV de ejemplo",
+	csvFileLabel: "Archivo CSV",
+	selectedFile: "Seleccionado: {name} ({size} KB)",
+	csvRequired: "Selecciona un archivo CSV",
+	uploadedToast: "{count, plural, one {Se subió # par de pregunta y respuesta} other {Se subieron # pares de pregunta y respuesta}}",
+	uploadErrorToast: "No se pudo subir el CSV",
+	upload: "Subir"
+};
 var platformSidebar$1 = {
 	collapseSidebar: "Contraer barra lateral",
 	expandSidebar: "Expandir barra lateral",
@@ -75738,6 +76398,12 @@ var es = {
 	workflowsDeleteWorkflowModal: workflowsDeleteWorkflowModal$1,
 	workflowsToolDialogs: workflowsToolDialogs$1,
 	ltiTabLabels: ltiTabLabels$1,
+	evaluationTabLabels: evaluationTabLabels$1,
+	benchmarksIndex: benchmarksIndex$1,
+	benchmarksCreateBenchmarkModal: benchmarksCreateBenchmarkModal$1,
+	benchmarksBenchmarkItemsModal: benchmarksBenchmarkItemsModal$1,
+	benchmarksDeleteItemConfirm: benchmarksDeleteItemConfirm$1,
+	benchmarksAddItemsModal: benchmarksAddItemsModal$1,
 	platformSidebar: platformSidebar$1,
 	platformNavbar: platformNavbar$1
 };
@@ -78856,7 +79522,9 @@ var profileAccount = {
 	descAdvanced: "配置组织的高级设置。",
 	descBasic: "管理您的基本账户信息和偏好设置。",
 	descSocial: "连接和管理您的社交媒体账户。",
-	descSecurity: "更新您的安全设置和密码。"
+	descSecurity: "更新您的安全设置和密码。",
+	tabBenchmarks: "基准测试",
+	descBenchmarks: "构建包含测试输入的基准测试套件，以评估智能体。"
 };
 var profileAddCreditsModal = {
 	title: "添加积分",
@@ -80858,6 +81526,220 @@ var ltiTabLabels = {
 		toolError: "LTI 工具出现问题"
 	}
 };
+var evaluationTabLabels = {
+	header: {
+		title: "评估",
+		description: "让此智能体在基准测试上运行并为回答打分。",
+		infoBox: "衡量智能体的实际表现。让它在基准测试（一组测试问题）上运行，每个回答都会被打分，帮助你发现薄弱环节，并在不断优化智能体的过程中追踪改进。"
+	},
+	benchmarkPicker: {
+		placeholder: "选择基准测试",
+		searchPlaceholder: "搜索基准测试...",
+		empty: "未找到基准测试"
+	},
+	actions: {
+		newExperiment: "新建评估",
+		manageBenchmarks: "管理基准测试",
+		viewResults: "查看结果",
+		evaluate: "新建审查",
+		exportCsv: "导出 CSV",
+		exporting: "导出中...",
+		checkStatus: "检查状态",
+		checkingStatus: "检查中...",
+		notReadyHint: "评估完成后可用",
+		"delete": "删除",
+		rowMenuAriaLabel: "{name} 的操作"
+	},
+	checkStatusToast: {
+		pending: "评估仍在运行",
+		completed: "评估已完成",
+		failed: "评估失败"
+	},
+	table: {
+		experimentColumn: "评估",
+		statusColumn: "状态",
+		initiatedByColumn: "发起人",
+		createdColumn: "创建时间",
+		actionsColumn: "操作",
+		emptyState: "此智能体在基准测试\"{dataset}\"上暂无评估。"
+	},
+	emptyBenchmarks: "还没有基准测试。请使用上方的\"管理基准测试\"创建一个。",
+	exportErrorToast: "导出评估失败",
+	startExperimentModal: {
+		title: "新建评估",
+		description: "让此智能体对所选基准测试中的每个条目运行。后台运行完成后，结果将显示在这里。",
+		benchmarkLabel: "基准测试",
+		experimentNameLabel: "评估名称",
+		experimentNamePlaceholder: "留空则自动生成",
+		cancel: "取消",
+		start: "开始评估",
+		successToast: "评估已开始",
+		errorToast: "启动评估失败"
+	},
+	manageBenchmarksModal: {
+		title: "管理基准测试",
+		description: "无需离开智能体设置即可创建和编辑租户级基准测试。"
+	},
+	detailModal: {
+		benchmarkLabel: "基准测试",
+		startedPrefix: "开始于",
+		tracesSuffixSingular: "条追踪已生成",
+		tracesSuffixPlural: "条追踪已生成",
+		waitingHint: "正在等待评估生成追踪。此页面会自动刷新。",
+		noScoresHint: "暂无评分",
+		newReview: "新建审查",
+		scoreNameLabel: "评分名称",
+		scoreNamePlaceholder: "准确率",
+		scoreValueLabel: "数值",
+		scoreValuePlaceholder: "0.0",
+		scoreCommentLabel: "备注",
+		scoreCommentPlaceholder: "可选",
+		addScore: "添加",
+		scoreInputError: "需要评分名称和数值",
+		scoreRangeError: "评分数值必须介于 0 和 1 之间",
+		scoreAddedToast: "评分已添加",
+		scoreErrorToast: "添加评分失败",
+		scoreRemovedToast: "评分已删除",
+		scoreRemoveErrorToast: "删除评分失败",
+		traceLabel: "追踪",
+		itemLabel: "条目",
+		removeScoreAriaLabel: "删除评分",
+		inputLabel: "输入",
+		expectedOutputLabel: "预期输出",
+		actualOutputLabel: "实际输出",
+		errorLabel: "错误",
+		noOutputHint: "暂无输出",
+		scoresHeading: "评分",
+		addScoreHeading: "添加评分",
+		reviewsHeading: "审查",
+		reviewsHint: "审查是以 LLM 作为评审的运行，会为此实验中的每条追踪打分。已完成审查的评分会显示在下方每个条目下。",
+		reviewsRunningLabel: "运行中",
+		reviewsCompletedLabel: "已完成",
+		reviewsFailedLabel: "失败",
+		reviewUnnamedScore: "未命名评分",
+		reviewFinishedPrefix: "完成于",
+		reviewStartedPrefix: "开始于",
+		reviewStatus: {
+			pending: "排队中",
+			running: "运行中",
+			completed: "已完成",
+			failed: "失败"
+		},
+		refresh: "刷新",
+		reviewDetailsAriaLabel: "切换审查详情",
+		reviewCriteriaLabel: "标准",
+		reviewScoresProducedLabel: "已生成的评分",
+		reviewNoScoresYet: "该审查暂无评分记录。",
+		scoresSuffixSingular: "项评分",
+		scoresSuffixPlural: "项评分"
+	},
+	evaluateModal: {
+		title: "评估",
+		description: "使用 LLM 为此评估中的每条追踪打分。",
+		criteriaLabel: "评判标准",
+		criteriaPlaceholder: "例如：从以下方面评估回答：\n1. 准确性\n2. 完整性\n3. 清晰度\n准确性权重最高。",
+		scoreNameLabel: "评分名称",
+		scoreNamePlaceholder: "例如 quality",
+		maxConcurrencyLabel: "最大并发数",
+		llmLabel: "LLM",
+		selectLlmPlaceholder: "选择 LLM",
+		cancel: "取消",
+		submit: "评估",
+		validationError: "需要填写评判标准、评分名称和 LLM",
+		successToast: "评估已开始",
+		errorToast: "启动评估失败",
+		pickerTitle: "选择 LLM",
+		pickerDescription: "先选择提供商，再选择模型。",
+		pickerSelectorAriaLabel: "LLM 模型选择器",
+		pickerSearchPlaceholder: "搜索提供商...",
+		pickerEmpty: "暂无提供商",
+		pickerLogoAlt: "{name} 徽标"
+	},
+	deleteModal: {
+		title: "删除评估",
+		confirmation: "确定要删除此评估吗？此操作无法撤销。",
+		cancel: "取消",
+		confirm: "删除",
+		confirming: "删除中...",
+		successToast: "评估已删除",
+		errorToast: "删除评估失败"
+	}
+};
+var benchmarksIndex = {
+	searchPlaceholder: "搜索基准测试",
+	searchAriaLabel: "搜索基准测试",
+	newBenchmark: "新建基准测试",
+	nameColumn: "名称",
+	descriptionColumn: "描述",
+	createdColumn: "创建时间",
+	actionsColumn: "操作",
+	viewItemsAriaLabel: "查看 {name} 中的条目",
+	loading: "加载中...",
+	empty: "未找到基准测试"
+};
+var benchmarksCreateBenchmarkModal = {
+	title: "新建基准测试",
+	description: "创建基准测试以归组用于智能体评估运行的测试输入。",
+	nameLabel: "名称",
+	namePlaceholder: "qa-eval-v1",
+	descriptionLabel: "描述",
+	descriptionPlaceholder: "这个基准测试的用途是什么？",
+	nameRequired: "基准测试名称为必填项",
+	cancel: "取消",
+	create: "创建",
+	successToast: "基准测试已创建",
+	errorToast: "创建基准测试失败"
+};
+var benchmarksBenchmarkItemsModal = {
+	description: "用于在此基准测试上评估智能体的问答对。",
+	addItems: "添加问答",
+	questionColumn: "问题",
+	answerColumn: "答案",
+	actionsColumn: "操作",
+	deleteItemAriaLabel: "删除问答",
+	empty: "此基准测试还没有问答"
+};
+var benchmarksDeleteItemConfirm = {
+	title: "删除问答",
+	confirmation: "确定要删除这组问答吗？此操作无法撤销。",
+	cancel: "取消",
+	"delete": "删除",
+	deleting: "删除中...",
+	successToast: "问答已删除",
+	errorToast: "删除问答失败"
+};
+var benchmarksAddItemsModal = {
+	title: "向 {name} 添加问答",
+	description: "直接添加问答对或上传 CSV 文件。",
+	manualTab: "手动",
+	csvTab: "CSV 上传",
+	questionLabel: "问题",
+	questionPlaceholder: "什么是机器学习？",
+	answerLabel: "答案",
+	answerPlaceholder: "可选",
+	removeRowAriaLabel: "删除行",
+	addRow: "添加行",
+	cancel: "取消",
+	add: "添加",
+	rowRequired: "至少添加一行并填写问题",
+	addedToast: "已添加 {count} 组问答",
+	addErrorToast: "添加问答失败",
+	csvFormatTitle: "CSV 格式",
+	csvRuleEncoding: "UTF-8 编码",
+	csvRuleRequiredPrefix: "必需列：",
+	csvRuleRequiredSuffix: "（问题）",
+	csvRuleOptionalPrefix: "可选列：",
+	csvRuleOptionalSuffix: "（答案）",
+	csvRuleMaxSize: "文件最大 10 MB，最多 10,000 行",
+	csvRuleBlankSkipped: "问题为空的行将被跳过",
+	downloadSample: "下载示例 CSV",
+	csvFileLabel: "CSV 文件",
+	selectedFile: "已选择：{name}（{size} KB）",
+	csvRequired: "请选择一个 CSV 文件",
+	uploadedToast: "已上传 {count} 组问答",
+	uploadErrorToast: "上传 CSV 失败",
+	upload: "上传"
+};
 var platformSidebar = {
 	collapseSidebar: "收起侧边栏",
 	expandSidebar: "展开侧边栏",
@@ -81120,6 +82002,12 @@ var zh = {
 	workflowsDeleteWorkflowModal: workflowsDeleteWorkflowModal,
 	workflowsToolDialogs: workflowsToolDialogs,
 	ltiTabLabels: ltiTabLabels,
+	evaluationTabLabels: evaluationTabLabels,
+	benchmarksIndex: benchmarksIndex,
+	benchmarksCreateBenchmarkModal: benchmarksCreateBenchmarkModal,
+	benchmarksBenchmarkItemsModal: benchmarksBenchmarkItemsModal,
+	benchmarksDeleteItemConfirm: benchmarksDeleteItemConfirm,
+	benchmarksAddItemsModal: benchmarksAddItemsModal,
 	platformSidebar: platformSidebar,
 	platformNavbar: platformNavbar
 };