fix(agents/failover): classify undici terminated and codex Request failed as timeout (#69368)

CHANGELOG.md

@@ -58,6 +58,8 @@ Docs: https://docs.openclaw.ai
 
 - Thinking defaults/status: raise the implicit default thinking level for reasoning-capable models from legacy `off`/`low` fallback behavior to a safe provider-supported `medium` equivalent when no explicit config default is set, preserve configured-model reasoning metadata when runtime catalog loading is empty, and make `/status` report the same resolved default as runtime.
 - Gateway/model pricing: fetch OpenRouter and LiteLLM pricing asynchronously at startup and extend catalog fetch timeouts to 30 seconds, reducing noisy timeout warnings during slow upstream responses.
+- Agents/failover: classify bare undici transport failures (`terminated`, `UND_ERR_SOCKET`, `UND_ERR_CONNECT_TIMEOUT`, body/header timeouts, aborted streams) and pi-ai's openai-codex `Request failed` sentinel as `timeout`, so Cloudflare 502s with empty bodies and mid-response socket resets actually enter the configured fallback chain instead of surfacing as unclassified errors. Fixes #69368. (#69677) Thanks @sk7n4k3d.
+- Providers/Anthropic Vertex: restore ADC-backed model discovery after the lightweight provider-discovery path by resolving emitted discovery entries, exposing synthetic auth on bootstrap discovery, and honoring copied env snapshots when probing the default GCP ADC path. Fixes #65715. (#65716) Thanks @feiskyer.
 - Plugins/install: add newly installed plugin ids to an existing `plugins.allow` list before enabling them, so allowlisted configs load installed plugins after restart.
 - Status: show `Fast` in `/status` when fast mode is enabled, including config/default-derived fast mode, and omit it when disabled.
 - OpenAI/image generation: detect Azure OpenAI-style image endpoints, use Azure `api-key` auth plus deployment-scoped image URLs, and honor `AZURE_OPENAI_API_VERSION` so image generation and edits work against Azure-hosted OpenAI resources. (#70570) Thanks @zhanggpcsu.

src/agents/pi-embedded-helpers.isbillingerrormessage.test.ts

@@ -844,6 +844,34 @@ describe("isFailoverErrorMessage", () => {
     expect(classifyFailoverReason(INTERNAL_SERVER_ERROR_STATUS_WITH_500_SAMPLE)).toBe("timeout");
     expect(isFailoverErrorMessage(INTERNAL_SERVER_ERROR_STATUS_WITH_500_SAMPLE)).toBe(true);
   });
+
+  it("matches bare undici transport failures as timeout (#69368)", () => {
+    expectTimeoutFailoverSamples([
+      "terminated",
+      "Terminated",
+      "  terminated  ",
+      "UND_ERR_SOCKET",
+      "Error: UND_ERR_SOCKET other side closed",
+      "UND_ERR_CONNECT_TIMEOUT",
+      "UND_ERR_HEADERS_TIMEOUT",
+      "UND_ERR_BODY_TIMEOUT",
+      "UND_ERR_ABORTED",
+      "UND_ERR_REQ_CONTENT_LENGTH_MISMATCH",
+    ]);
+  });
+
+  it("matches pi-ai openai-codex bare transport failures as timeout (#69368)", () => {
+    expectTimeoutFailoverSamples([
+      "Request failed",
+      "request failed",
+      "  Request failed  ",
+      "Request failed after repeated internal retries.",
+    ]);
+  });
+
+  it("does not classify unrelated 'terminated' prose as timeout", () => {
+    expectNotFailoverSample("The user terminated the session manually.");
+  });
 });
 
 describe("parseImageSizeError", () => {

src/agents/pi-embedded-helpers/failover-matches.ts

@@ -126,6 +126,18 @@ const ERROR_PATTERNS = {
     // falls through to reason=unknown (#58315).
     /\boperation was aborted\b/i,
     /\bstream (?:was )?(?:closed|aborted)\b/i,
+    // Undici transport-level failures during CDN/provider outages (Cloudflare
+    // 502 served with an empty body, socket reset mid-response, body-stream
+    // aborted). These arrive as bare strings on the outer error and, without
+    // an explicit match, the fallback chain is never attempted (#69368).
+    /^terminated$/i,
+    /\bund_err_(?:socket|connect|headers?|body|req_content_length_mismatch|aborted|closed)\b/i,
+    // pi-ai's openai-codex provider surfaces `Request failed` when the HTTP
+    // response has no body and no status text (typical of Cloudflare 502s
+    // from the upstream Codex service). Treat it as a transport failure so
+    // the configured fallback chain runs instead of surfacing the error.
+    /^request failed$/i,
+    /\brequest failed after repeated internal retries\b/i,
   ],
   billing: [
     /["']?(?:status|code)["']?\s*[:=]\s*402\b|\bhttp\s*402\b|\berror(?:\s+code)?\s*[:=]?\s*402\b|\b(?:got|returned|received)\s+(?:a\s+)?402\b|^\s*402\s+payment/i,