fix(inference): hide run picker when the scenario has no runs on the date

Follow-up to the scenario-scoped run picker: the empty-coverage fallback
in restrictRunsToScenario() fired whenever scenarioRunIds was empty,
which conflated two cases — "no per-run coverage rows for the date"
(old data, where falling back is right) and "coverage exists but zero
runs for this scenario" (e.g. 2026-07-02 / 2026-07-04, where only
single_turn sweeps ran). In the second case the picker degraded to the
old leaky behavior and listed 6 / 3 single_turn runs under Agentic
Traces.

restrictRunsToScenario() now takes coverageKnown
(runConfigs.length > 0): when coverage data is unavailable it still
falls back to the changelog-scoped list, but when coverage exists and
the scenario has no runs it returns null so the picker hides — matching
the (empty) chart instead of offering meaningless "as of run" cutoffs.

中文：场景化运行选择器的后续修复：restrictRunsToScenario() 原先只要
scenarioRunIds 为空就触发回退，混淆了两种情况——"当天没有任何按运行
的覆盖数据行"（旧数据，回退是正确的）与"覆盖数据存在但该场景当天没有
任何运行"（如 2026-07-02 / 2026-07-04，当天只有 single_turn 运行）。
在第二种情况下选择器退化回旧的泄漏行为，在 Agentic Traces 场景下列出
6 / 3 个 single_turn 运行。现在 restrictRunsToScenario() 接收
coverageKnown（runConfigs.length > 0）：覆盖数据缺失时仍回退到
changelog 范围列表；覆盖数据存在但该场景无运行时返回 null，选择器随之
隐藏——与（空的）图表保持一致，而不是提供无意义的 "as of run" 截断点。

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

Author: cquil11

packages/app/src/components/inference/InferenceContext.tsx

@@ -395,12 +395,21 @@ export function InferenceProvider({
   );
 
   // The picker list: the changelog/precision-scoped runs, further restricted to
-  // the ones that shipped data for the selected scenario. Falls back to the
-  // changelog-scoped list when coverage data is unavailable (old dates / JSON
-  // mode) so the selector always renders.
+  // the ones that shipped data for the selected scenario. When the date has
+  // coverage rows but none for this scenario (e.g. only single_turn sweeps ran
+  // that day while viewing Agentic Traces) the picker hides (null) rather than
+  // listing other-scenario runs; only when coverage data is unavailable
+  // altogether (old dates / JSON snapshots) does it fall back to the
+  // changelog-scoped list so the selector still renders.
   const filteredAvailableRuns = useMemo(
-    () => restrictRunsToScenario(changelogFilteredRuns, availableRuns, scenarioRunIds),
-    [changelogFilteredRuns, availableRuns, scenarioRunIds],
+    () =>
+      restrictRunsToScenario(
+        changelogFilteredRuns,
+        availableRuns,
+        scenarioRunIds,
+        runConfigs.length > 0,
+      ),
+    [changelogFilteredRuns, availableRuns, scenarioRunIds, runConfigs],
   );
 
   const effectiveSelectedRunId = useMemo(() => {

packages/app/src/lib/utils.test.ts

@@ -550,38 +550,50 @@ describe('restrictRunsToScenario', () => {
   };
 
   it('keeps only runs present in the scenario coverage set', () => {
-    const result = restrictRunsToScenario(runs, runs, new Set(['10']));
+    const result = restrictRunsToScenario(runs, runs, new Set(['10']), true);
     expect(Object.keys(result!)).toEqual(['10']);
   });
 
-  it('falls back to the changelog-scoped list when coverage is empty', () => {
-    const result = restrictRunsToScenario(runs, runs, new Set());
+  it('falls back to the changelog-scoped list when coverage data is unavailable', () => {
+    // Old dates / JSON snapshots without per-run coverage rows: we can't say
+    // anything about scenarios, so the picker keeps the changelog-scoped list.
+    const result = restrictRunsToScenario(runs, runs, new Set(), false);
     expect(result).toBe(runs);
   });
 
+  it('hides the picker when coverage exists but the scenario has no runs that date', () => {
+    // Repro: 2026-07-02 / 2026-07-04 had only single_turn runs; on the Agentic
+    // Traces scenario the picker must show NO runs (null), not fall back to
+    // listing the single_turn runs.
+    const result = restrictRunsToScenario(runs, runs, new Set(), true);
+    expect(result).toBeNull();
+  });
+
   it('pulls a scenario run from the raw map when the changelog list omits it', () => {
     // A run that shipped data without a changelog entry: filterRunsByModel drops
     // it, but it is still a valid scenario run the picker must show.
     const changelogFiltered = { '10': runs['10'] };
     const allRuns = runs;
-    const result = restrictRunsToScenario(changelogFiltered, allRuns, new Set(['10', '20']));
+    const result = restrictRunsToScenario(changelogFiltered, allRuns, new Set(['10', '20']), true);
     expect(Object.keys(result!).toSorted()).toEqual(['10', '20']);
     expect(result!['20']).toBe(runs['20']);
   });
 
   it('prefers the changelog-scoped entry over the raw one', () => {
     const scoped = makeRun([['dsv4-fp4-b200-sglang']], { runId: '10', conclusion: 'scoped' });
-    const result = restrictRunsToScenario({ '10': scoped }, runs, new Set(['10']));
+    const result = restrictRunsToScenario({ '10': scoped }, runs, new Set(['10']), true);
     expect(result!['10']).toBe(scoped);
   });
 
-  it('falls back to the changelog list when no scenario run survives', () => {
-    const result = restrictRunsToScenario(runs, runs, new Set(['999']));
-    expect(result).toBe(runs);
+  it('returns null when no scenario run survives the restriction', () => {
+    // Scenario runs exist in coverage but none resolve to a listable RunInfo
+    // (e.g. still in progress, so absent from the workflow-runs map).
+    const result = restrictRunsToScenario(runs, runs, new Set(['999']), true);
+    expect(result).toBeNull();
   });
 
-  it('returns null changelog list unchanged', () => {
-    expect(restrictRunsToScenario(null, null, new Set(['10']))).toBeNull();
+  it('returns null changelog list unchanged when coverage is unavailable', () => {
+    expect(restrictRunsToScenario(null, null, new Set(), false)).toBeNull();
   });
 });
 

packages/app/src/lib/utils.ts

@@ -300,25 +300,33 @@ export function filterRunsByModel(
  * entry, so `filterRunsByModel` may omit it while it is still a legitimate
  * scenario run the user must be able to select.
  *
- * Fallbacks that keep the selector rendering:
- *  - `scenarioRunIds` empty → no coverage data (dates predating per-run
- *    coverage, or JSON/fixtures mode): return the changelog-scoped list
- *    unchanged rather than blanking the picker.
- *  - Non-empty coverage but zero surviving runs (shouldn't happen in practice)
- *    → also fall back to the changelog-scoped list.
+ * `coverageKnown` says whether per-(run, config) coverage rows exist for the
+ * date at all (`runConfigs.length > 0`). It distinguishes two very different
+ * "empty `scenarioRunIds`" cases:
+ *  - `coverageKnown === false` → no coverage data (dates predating per-run
+ *    coverage, or JSON/fixtures snapshots without it): we can't say anything
+ *    about scenarios, so return the changelog-scoped list unchanged rather
+ *    than blanking the picker.
+ *  - `coverageKnown === true` and no scenario run survives → the date has
+ *    benchmark data but NONE for this scenario (e.g. only single_turn sweeps
+ *    ran that day while the user views Agentic Traces). Return null so the
+ *    picker hides instead of listing other-scenario runs — showing them is
+ *    exactly the leak this helper exists to prevent, and selecting one would
+ *    constrain an already-empty chart to a meaningless "as of run" cutoff.
  */
 export function restrictRunsToScenario(
   changelogFiltered: Record<string, RunInfo> | null,
   allRuns: Record<string, RunInfo> | null,
   scenarioRunIds: Set<string>,
+  coverageKnown: boolean,
 ): Record<string, RunInfo> | null {
-  if (scenarioRunIds.size === 0) return changelogFiltered;
+  if (!coverageKnown) return changelogFiltered;
   const restricted: Record<string, RunInfo> = {};
   for (const runId of scenarioRunIds) {
     const info = changelogFiltered?.[runId] ?? allRuns?.[runId];
     if (info) restricted[runId] = info;
   }
-  return Object.keys(restricted).length > 0 ? restricted : changelogFiltered;
+  return Object.keys(restricted).length > 0 ? restricted : null;
 }
 
 /**