fix(dashboard): trust provider topology live state (#6322)

Trust live provider-topology state on the Home dashboard. Integrated into release/v3.8.46.

CHANGELOG.md

@@ -20,6 +20,7 @@
 
 ### 🐛 Bug Fixes
 
+- **fix(dashboard):** the Home provider-topology widget now trusts the live provider-metrics snapshot — it uses `topology.errorProvider` and live `activeRequests` directly instead of re-deriving state from a stale `lastErrorAt` or applying a frontend timeout filter, so the topology reflects real-time provider health. Regression guard: `tests/unit/home-provider-topology-live-state.test.ts`. (thanks @xz-dev)
 - **fix(sse):** strip zero-width markers from streamed **tool-call arguments** — a follow-up to [#5857](https://github.com/diegosouzapw/OmniRoute/pull/5857). That PR removed injected zero-width joiners (U+200D) from streamed assistant text/reasoning but deliberately left tool-call argument JSON byte-exact. The request-side obfuscation (`open-sse/services/claudeCodeObfuscation.ts`) injects ZWJ into agent words — including the temp path inside the Bash tool description — and Claude models copy that verbatim into generated commands, which are delivered as tool-call arguments rather than assistant text. As a result the ZWJ survived and corrupted code blocks (e.g. a temp path rendered with an invisible joiner). Now `open-sse/handlers/responseSanitizer.ts` strips zero-width code points from tool-call argument strings at every emit site (OpenAI non-stream/stream chat `tool_calls` + legacy `function_call`, native Responses `function_call` items, the OpenAI→Responses conversion, and the native Responses streaming `response.function_call_arguments.delta/.done` events). Only zero-width code points are removed; JSON structure and all other bytes stay identical (no parse/restringify), so normal arguments remain byte-exact. Regression guard: 6 new cases in `tests/unit/response-sanitizer.test.ts` (suite 50/50).
 
 - **fix(nodejs):** the default app log path now resolves under `DATA_DIR` (`~/.omniroute/logs/application/app.log`) instead of `process.cwd()` ([#6197](https://github.com/diegosouzapw/OmniRoute/issues/6197)) — the globally-installed CLI runs from an arbitrary working directory, so anchoring the default to cwd made file logging silently write to (or no-op under) an unrelated directory, contradicting the documented `.env.example` default. `getAppLogFilePath()` now computes the default lazily via the pure `resolveDataDir()` resolver (honours a per-process `DATA_DIR`, no directory-creation side effect); an explicit `APP_LOG_FILE_PATH` still wins. Regression guard: `tests/unit/logenv-datadir-path-6197.test.ts` (3).

src/app/(dashboard)/dashboard/HomePageClient.tsx

@@ -112,6 +112,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
   const [baseUrl, setBaseUrl] = useState("/v1");
   const [selectedProvider, setSelectedProvider] = useState(null);
   const [providerMetrics, setProviderMetrics] = useState<Record<string, ProviderMetricSummary>>({});
+  const [providerTopology, setProviderTopology] = useState({ lastProvider: "", errorProvider: "" });
   const [providerNodes, setProviderNodes] = useState<
     Array<{ id?: string; prefix?: string; name?: string }>
   >([]);
@@ -304,6 +305,10 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
           const data = await metricsRes.json();
           if (!cancelled) {
             setProviderMetrics(data.metrics || {});
+            setProviderTopology({
+              lastProvider: normalizeProviderId(data.topology?.lastProvider),
+              errorProvider: normalizeProviderId(data.topology?.errorProvider),
+            });
           }
         }
       } catch (error) {
@@ -494,28 +499,7 @@ export default function HomePageClient({ machineId }: HomePageClientProps) {
     return Array.from(byProvider.values());
   }, [providerStats, providerMetrics, providerNodes]);
 
-  const { lastProvider, errorProvider } = useMemo(() => {
-    let recentProvider = "";
-    let recentTimestamp = 0;
-    let recentErrorProvider = "";
-    let recentErrorTimestamp = 0;
-
-    for (const [provider, metrics] of Object.entries(providerMetrics)) {
-      const requestTimestamp = metrics.lastRequestAt ? Date.parse(metrics.lastRequestAt) : 0;
-      if (Number.isFinite(requestTimestamp) && requestTimestamp > recentTimestamp) {
-        recentProvider = normalizeProviderId(provider);
-        recentTimestamp = requestTimestamp;
-      }
-
-      const errorTimestamp = metrics.lastErrorAt ? Date.parse(metrics.lastErrorAt) : 0;
-      if (Number.isFinite(errorTimestamp) && errorTimestamp > recentErrorTimestamp) {
-        recentErrorProvider = normalizeProviderId(provider);
-        recentErrorTimestamp = errorTimestamp;
-      }
-    }
-
-    return { lastProvider: recentProvider, errorProvider: recentErrorProvider };
-  }, [providerMetrics]);
+  const { lastProvider, errorProvider } = providerTopology;
 
   const pollBackgroundUpdate = useCallback(
     async ({

src/app/(dashboard)/home/ProviderTopology.tsx

@@ -1,6 +1,6 @@
 "use client";
 
-import { useMemo, useState, useEffect, useRef } from "react";
+import { useMemo } from "react";
 import { useTranslations } from "next-intl";
 import { Handle, Position, type Node, type Edge, type NodeTypes } from "@xyflow/react";
 import { AI_PROVIDERS } from "@/shared/constants/providers";
@@ -10,9 +10,6 @@ import { StatusDot } from "@/shared/components/flow/StatusDot";
 import { edgeStyle } from "@/shared/components/flow/edgeStyles";
 import { resolveTopologyNodeLabel } from "./topologyLabel";
 
-const FE_ACTIVE_TIMEOUT_MS = 60_000;
-const FE_ACTIVE_TICK_MS = 1_000;
-
 // Rings: [capacity, rx, ry]. Each successive ring fits ~6 more nodes.
 const RINGS: [number, number, number][] = [
   [8, 210, 132],
@@ -271,44 +268,13 @@ export default function ProviderTopology({
   const lastKey = lastProvider.toLowerCase();
   const errorKey = errorProvider.toLowerCase();
 
-  const rawActiveSet = useMemo(
+  const activeSet = useMemo(
     () => new Set<string>(activeKey ? activeKey.split(",") : []),
     [activeKey]
   );
   const lastSet = useMemo(() => new Set<string>(lastKey ? [lastKey] : []), [lastKey]);
   const errorSet = useMemo(() => new Set<string>(errorKey ? [errorKey] : []), [errorKey]);
 
-  const firstSeenRef = useRef<Record<string, number>>({});
-  const [tick, setTick] = useState(0);
-
-  useEffect(() => {
-    const seen = firstSeenRef.current;
-    const now = Date.now();
-    for (const p of rawActiveSet) {
-      if (!seen[p]) seen[p] = now;
-    }
-    for (const p of Object.keys(seen)) {
-      if (!rawActiveSet.has(p)) delete seen[p];
-    }
-  }, [rawActiveSet]);
-
-  useEffect(() => {
-    if (rawActiveSet.size === 0) return;
-    const id = setInterval(() => setTick((t) => t + 1), FE_ACTIVE_TICK_MS);
-    return () => clearInterval(id);
-  }, [rawActiveSet]);
-
-  const activeSet = useMemo(() => {
-    const now = Date.now();
-    const filtered = new Set<string>();
-    for (const p of rawActiveSet) {
-      const ts = firstSeenRef.current[p];
-      if (!ts || now - ts < FE_ACTIVE_TIMEOUT_MS) filtered.add(p);
-    }
-    return filtered;
-    // eslint-disable-next-line react-hooks/exhaustive-deps
-  }, [rawActiveSet, tick]);
-
   const { nodes, edges } = useMemo(
     () => buildLayout(providers, activeSet, lastSet, errorSet),
     // eslint-disable-next-line react-hooks/exhaustive-deps

tests/unit/home-provider-topology-live-state.test.ts

@@ -0,0 +1,45 @@
+import test from "node:test";
+import assert from "node:assert/strict";
+import { readFileSync } from "node:fs";
+import { fileURLToPath } from "node:url";
+
+const homePageClientSrc = readFileSync(
+  fileURLToPath(new URL("../../src/app/(dashboard)/dashboard/HomePageClient.tsx", import.meta.url)),
+  "utf8"
+);
+
+const providerTopologySrc = readFileSync(
+  fileURLToPath(new URL("../../src/app/(dashboard)/home/ProviderTopology.tsx", import.meta.url)),
+  "utf8"
+);
+
+test("home topology uses provider-metrics topology.errorProvider instead of re-deriving from stale lastErrorAt", () => {
+  assert.match(
+    homePageClientSrc,
+    /errorProvider:\s*normalizeProviderId\(data\.topology\?\.errorProvider\)/,
+    "HomePageClient should trust /api/provider-metrics topology.errorProvider"
+  );
+
+  const localTopologyDerivation = homePageClientSrc.match(
+    /const \{ lastProvider, errorProvider \} = useMemo[\s\S]*?\}, \[providerMetrics\]\);/
+  );
+  assert.equal(
+    localTopologyDerivation,
+    null,
+    "HomePageClient must not re-derive topology error state from providerMetrics.lastErrorAt"
+  );
+});
+
+test("ProviderTopology treats live activeRequests as the current snapshot without frontend timeout filtering", () => {
+  assert.doesNotMatch(
+    providerTopologySrc,
+    /FE_ACTIVE_TIMEOUT_MS|FE_ACTIVE_TICK_MS|firstSeenRef|setInterval\(/,
+    "ProviderTopology must not expire long-running live requests on its own"
+  );
+
+  assert.match(
+    providerTopologySrc,
+    /const activeSet = useMemo\(\s*\(\) => new Set<string>\(activeKey \? activeKey\.split\(","\) : \[\]\),\s*\[activeKey\]\s*\);/,
+    "activeSet should be derived directly from activeRequests/current live snapshot"
+  );
+});