release/v0.59.2

api/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "api"
-version = "0.59.1"
+version = "0.59.2"
 description = "Agenta API"
 authors = [
     { name = "Mahmoud Mabrouk", email = "[email protected]" },

sdk/agenta/client/backend/client.py

@@ -91,7 +91,9 @@ def __init__(
         _defaulted_timeout = (
             timeout
             if timeout is not None
-            else 60 if httpx_client is None else httpx_client.timeout.read
+            else 60
+            if httpx_client is None
+            else httpx_client.timeout.read
         )
         self._client_wrapper = SyncClientWrapper(
             base_url=_get_base_url(base_url=base_url, environment=environment),
@@ -652,7 +654,9 @@ def __init__(
         _defaulted_timeout = (
             timeout
             if timeout is not None
-            else 60 if httpx_client is None else httpx_client.timeout.read
+            else 60
+            if httpx_client is None
+            else httpx_client.timeout.read
         )
         self._client_wrapper = AsyncClientWrapper(
             base_url=_get_base_url(base_url=base_url, environment=environment),

sdk/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "agenta"
-version = "0.59.1"
+version = "0.59.2"
 description = "The SDK for agenta is an open-source LLMOps platform."
 readme = "README.md"
 authors = [

web/ee/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@agenta/ee",
-    "version": "0.59.1",
+    "version": "0.59.2",
     "private": true,
     "engines": {
         "node": ">=18"

web/ee/src/components/EvalRunDetails/AutoEvalRun/assets/AutoEvalRunSkeleton.tsx

@@ -2,8 +2,8 @@ import {memo} from "react"
 
 import {useRouter} from "next/router"
 
+import EvalRunOverviewViewerSkeleton from "../../components/EvalRunOverviewViewer/assets/EvalRunOverviewViewerSkeleton"
 import EvalRunHeaderSkeleton from "../components/EvalRunHeader/assets/EvalRunHeaderSkeleton"
-import EvalRunOverviewViewerSkeleton from "../components/EvalRunOverviewViewer/assets/EvalRunOverviewViewerSkeleton"
 import EvalRunPromptConfigViewerSkeleton from "../components/EvalRunPromptConfigViewer/assets/EvalRunPromptConfigViewerSkeleton"
 import EvalRunTestCaseViewerSkeleton from "../components/EvalRunTestCaseViewer/assets/EvalRunTestCaseViewerSkeleton"
 
@@ -19,7 +19,7 @@ const AutoEvalRunSkeleton = () => {
             ) : viewType === "prompt" ? (
                 <EvalRunPromptConfigViewerSkeleton />
             ) : (
-                <EvalRunOverviewViewerSkeleton />
+                <EvalRunOverviewViewerSkeleton className="px-6" />
             )}
         </section>
     )

web/ee/src/components/EvalRunDetails/AutoEvalRun/components/EvalRunFocusDrawer/assets/FocusDrawerContent/assets/RunTraceHeader.tsx

@@ -7,10 +7,7 @@ import dynamic from "next/dynamic"
 import EvalNameTag from "@/oss/components/EvalRunDetails/AutoEvalRun/assets/EvalNameTag"
 import {EVAL_TAG_COLOR} from "@/oss/components/EvalRunDetails/AutoEvalRun/assets/utils"
 import {useRunId} from "@/oss/contexts/RunIdContext"
-import {
-    evalAtomStore,
-    evaluationRunStateFamily,
-} from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms"
+import {evaluationRunStateFamily} from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms"
 import {useInvocationResult} from "@/oss/lib/hooks/useInvocationResult"
 
 const GenerationResultUtils = dynamic(
@@ -35,8 +32,7 @@ const RunTraceHeader = ({
     showComparisons?: boolean
 }) => {
     const baseRunId = useRunId()
-    const store = evalAtomStore()
-    const state = useAtomValue(evaluationRunStateFamily(rId), {store})
+    const state = useAtomValue(evaluationRunStateFamily(rId))
     const enriched = state?.enrichedRun
     const {trace: runTrace} = useInvocationResult({
         scenarioId: scId,

web/ee/src/components/EvalRunDetails/AutoEvalRun/components/EvalRunFocusDrawer/assets/FocusDrawerContent/index.tsx

@@ -35,9 +35,8 @@ const failureRunTypes = [EvaluationStatus.FAILED, EvaluationStatus.FAILURE, Eval
 const EMPTY_COMPARISON_RUN_IDS: string[] = []
 
 const FocusDrawerContent = () => {
-    const appState = useAppState()
-    const store = evalAtomStore()
     const router = useRouter()
+    const appState = useAppState()
 
     const [windowHight, setWindowHight] = useState(0)
     const [activeKeys, setActiveKeys] = useState<(string | number)[]>([
@@ -92,7 +91,7 @@ const FocusDrawerContent = () => {
         () => comparisonRunsStepsAtom(comparisonRunIds),
         [comparisonRunIds],
     )
-    const comparisonRunsSteps = useAtomValue(comparisonRunsStepsAtomInstance, {store})
+    const comparisonRunsSteps = useAtomValue(comparisonRunsStepsAtomInstance)
     // // Derive whether to show comparison mode
     const showComparisons = useMemo(
         () => Boolean(isBaseRun && comparisonRunIds.length > 0),

web/ee/src/components/EvalRunDetails/AutoEvalRun/components/EvalRunScoreTable/index.tsx

@@ -410,7 +410,13 @@ const EvalRunScoreTable = ({className}: {className?: string}) => {
                     {"h-[calc(100%-60px)]": !isComparison},
                 ])}
             >
-                <div className={clsx(["overflow-y-auto", {"w-[50%]": !isComparison}])}>
+                <div
+                    className={clsx([
+                        "overflow-y-auto",
+                        {"w-[50%]": !isComparison},
+                        {"w-full": chartMetrics.length < 3},
+                    ])}
+                >
                     <Table
                         dataSource={dataSource}
                         columns={useMemo(() => {
@@ -484,24 +490,26 @@ const EvalRunScoreTable = ({className}: {className?: string}) => {
                         rowKey={(r) => r.key}
                     />
                 </div>
-                <EvaluatorMetricsSpiderChart
-                    className={clsx([
-                        "min-h-[400px] h-[400px]",
-                        {"w-[50%] !h-full": !isComparison},
-                    ])}
-                    metrics={chartMetrics}
-                    series={useMemo(() => {
-                        return allRunIds.map((id, idx) => {
-                            const state = evalById[id]
-                            const compareIdx = state?.compareIndex || idx + 1
-                            return {
-                                key: idx === 0 ? "value" : `value-${idx + 1}`,
-                                color: (EVAL_COLOR as any)[compareIdx] || "#3B82F6",
-                                name: state?.enrichedRun?.name || `Eval ${compareIdx}`,
-                            }
-                        })
-                    }, [allRunIds, evalById])}
-                />
+                {chartMetrics.length < 3 ? null : (
+                    <EvaluatorMetricsSpiderChart
+                        className={clsx([
+                            "min-h-[400px] h-[400px]",
+                            {"w-[50%] !h-full": !isComparison},
+                        ])}
+                        metrics={chartMetrics}
+                        series={useMemo(() => {
+                            return allRunIds.map((id, idx) => {
+                                const state = evalById[id]
+                                const compareIdx = state?.compareIndex || idx + 1
+                                return {
+                                    key: idx === 0 ? "value" : `value-${idx + 1}`,
+                                    color: (EVAL_COLOR as any)[compareIdx] || "#3B82F6",
+                                    name: state?.enrichedRun?.name || `Eval ${compareIdx}`,
+                                }
+                            })
+                        }, [allRunIds, evalById])}
+                    />
+                )}
             </div>
         </div>
     )

web/ee/src/components/EvalRunDetails/AutoEvalRun/components/EvaluatorMetricsChart/index.tsx

@@ -170,10 +170,8 @@ const EvaluatorMetricsChart = ({
         }
         return []
     }, [metric, isBooleanMetric, hasDistribution, isNumeric])
-    console.log("chartData", chartData)
 
     const showHistogram = !isCompare || selectedItem === "histogram"
-    const showAverageBars = isCompare && selectedItem === "average"
 
     const formatYAxisTick = useCallback(
         (value: number) => {

web/ee/src/components/EvalRunDetails/AutoEvalRun/index.tsx

@@ -10,7 +10,7 @@ import AutoEvalRunSkeleton from "./assets/AutoEvalRunSkeleton"
 import {AutoEvalRunDetailsProps} from "./assets/types"
 import EvalRunHeader from "./components/EvalRunHeader"
 
-const EvalRunOverviewViewer = dynamic(() => import("./components/EvalRunOverviewViewer"), {
+const EvalRunOverviewViewer = dynamic(() => import("../components/EvalRunOverviewViewer"), {
     ssr: false,
 })
 const EvalRunPromptConfigViewer = dynamic(() => import("./components/EvalRunPromptConfigViewer"), {

web/ee/src/components/EvalRunDetails/HumanEvalRun/components/AnnotateScenarioButton/index.tsx

@@ -4,7 +4,6 @@ import {Button} from "antd"
 import {useAtomValue} from "jotai"
 
 import {AnnotationDto} from "@/oss/lib/hooks/useAnnotations/types"
-import {evalAtomStore} from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms"
 import {scenarioUiFlagsFamily} from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms/progress"
 import {getProjectValues} from "@/oss/state/project"
 
@@ -27,8 +26,7 @@ const AnnotateScenarioButton = ({
     className,
 }: AnnotateScenarioButtonProps) => {
     const [annotating, setAnnotating] = useState(false)
-    const store = evalAtomStore()
-    const uiFlags = useAtomValue(scenarioUiFlagsFamily({scenarioId, runId}), {store})
+    const uiFlags = useAtomValue(scenarioUiFlagsFamily({scenarioId, runId}))
     const isLoading = annotating || uiFlags.isAnnotating || uiFlags.isRevalidating
 
     const onAnnotate = useCallback(async () => {

web/ee/src/components/EvalRunDetails/HumanEvalRun/components/EvalResultsView/index.tsx

@@ -1,12 +1,5 @@
-import {memo, useCallback, useMemo} from "react"
-
-import deepEqual from "fast-deep-equal"
-import {useAtomValue} from "jotai"
-import {selectAtom} from "jotai/utils"
-
-import {evaluationEvaluatorsFamily} from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms"
-
-import EvaluatorMetricsCard from "./EvaluatorMetricsCard"
+import {memo} from "react"
+import EvalRunOverviewViewer from "../../../components/EvalRunOverviewViewer"
 
 /**
  * Displays run-level evaluation results grouped by evaluator.
@@ -15,23 +8,9 @@ import EvaluatorMetricsCard from "./EvaluatorMetricsCard"
  * are handled inside each card.
  */
 const EvalResultsView = ({runId}: {runId: string}) => {
-    const slugSelector = useCallback(
-        (list: any[] | undefined): string[] =>
-            (list || []).map((ev) => ev.slug || ev.id || ev.name),
-        [],
-    )
-
-    const slugsAtom = useMemo(
-        () => selectAtom(evaluationEvaluatorsFamily(runId), slugSelector, deepEqual),
-        [runId],
-    )
-    const evaluatorSlugs = useAtomValue(slugsAtom)
-
     return (
-        <section className="flex flex-wrap gap-2 overflow-y-auto p-1">
-            {evaluatorSlugs.map((slug) => (
-                <EvaluatorMetricsCard key={slug} runId={runId} evaluatorSlug={slug} />
-            ))}
+        <section className="overflow-y-auto flex flex-col gap-4">
+            <EvalRunOverviewViewer />
         </section>
     )
 }

web/ee/src/components/EvalRunDetails/HumanEvalRun/components/EvalRunBatchActions.tsx

@@ -1,7 +1,7 @@
 import {memo, useCallback, useState} from "react"
 
 import RunButton from "@agenta/oss/src/components/Playground/assets/RunButton"
-import {useAtomValue} from "jotai"
+import {getDefaultStore, useAtomValue} from "jotai"
 import {loadable} from "jotai/utils"
 
 // agenta hooks & utils
@@ -11,7 +11,6 @@ import {useEvalScenarioQueue} from "@/oss/lib/hooks/useEvalScenarioQueue"
 import {
     scenarioStepFamily,
     scenariosFamily,
-    evalAtomStore,
 } from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms"
 import {scenarioMetricsMapFamily} from "@/oss/lib/hooks/useEvaluationRunData/assets/atoms/runScopedMetrics"
 
@@ -35,19 +34,18 @@ const EMPTY_ROWS: any[] = []
 const EvalRunBatchActions = ({name}: {name: string}) => {
     const [rows, setRows] = useState<any[]>(EMPTY_ROWS)
     const runId = useRunId()
-    const store = evalAtomStore()
 
     const {enqueueScenario} = useEvalScenarioQueue({concurrency: 5, runId})
 
     // Lightweight subscription: only track the count of runnable scenarios - use global store
-    const hasRunnable = useAtomValue(hasRunnableScenarioFamily(runId), {store})
+    const hasRunnable = useAtomValue(hasRunnableScenarioFamily(runId))
     const isRunAllDisabled = !hasRunnable
 
     const handleRunAll = useCallback(async () => {
         if (!runId) return
 
         try {
-            const store = evalAtomStore()
+            const store = getDefaultStore()
 
             // Get all scenarios for this run (same as single run approach)
             const scenarios = store.get(scenariosFamily(runId))
@@ -132,7 +130,7 @@ const EvalRunBatchActions = ({name}: {name: string}) => {
         if (!runId) return []
 
         // 1. Gather the scenario IDs present in the current evaluation (sync)
-        const store = evalAtomStore()
+        const store = getDefaultStore()
         const scenarios = store.get(scenariosFamily(runId))
         const ids = scenarios.map((s: any) => s.id)