diff --git a/actions/setup/js/send_otlp_span.cjs b/actions/setup/js/send_otlp_span.cjs
index a974d1fcdc9..99e916a7177 100644
--- a/actions/setup/js/send_otlp_span.cjs
+++ b/actions/setup/js/send_otlp_span.cjs
@@ -1648,7 +1648,7 @@ async function sendJobConclusionSpan(spanName, options = {}) {
   const isAgentCancelled = agentConclusion === "cancelled";
   const isAgentNonOK = isAgentFailure || isAgentCancelled;
   // STATUS_CODE_ERROR = 2, STATUS_CODE_OK = 1
-  const statusCode = isAgentNonOK ? 2 : 1;
+  let statusCode = isAgentNonOK ? 2 : 1;
   let statusMessage;
   if (isAgentFailure) {
     statusMessage = `agent ${agentConclusion}`;
@@ -1675,6 +1675,15 @@ async function sendJobConclusionSpan(spanName, options = {}) {
     runStatus = "failure";
   }
 
+  // When GH_AW_AGENT_CONCLUSION and workflowRunConclusion are both absent (e.g. in the
+  // agent job's own post-step where needs.<job>.result is not yet visible), fall back to
+  // observable failure evidence so gh-aw.run.status and status.code are accurate.
+  if (!rawRunStatus && outputErrors.length > 0) {
+    runStatus = "failure";
+    statusCode = 2;
+    statusMessage = (errorMessages.length > 0 ? `errors detected: ${errorMessages[0]}` : "errors detected").slice(0, 256);
+  }
+
   if (isAgentFailure && errorMessages.length > 0) {
     statusMessage = `agent ${agentConclusion}: ${errorMessages[0]}`.slice(0, 256);
   }
diff --git a/actions/setup/js/send_otlp_span.test.cjs b/actions/setup/js/send_otlp_span.test.cjs
index 1d448e37c22..a984a43bd1c 100644
--- a/actions/setup/js/send_otlp_span.test.cjs
+++ b/actions/setup/js/send_otlp_span.test.cjs
@@ -4292,6 +4292,110 @@ describe("sendJobConclusionSpan", () => {
     });
   });
 
+  describe("run.status fallback from observable error signals", () => {
+    let readFileSpy;
+
+    beforeEach(() => {
+      readFileSpy = vi.spyOn(fs, "readFileSync").mockImplementation(() => {
+        throw Object.assign(new Error("ENOENT"), { code: "ENOENT" });
+      });
+    });
+
+    afterEach(() => {
+      readFileSpy.mockRestore();
+    });
+
+    it("sets gh-aw.run.status=failure and STATUS_CODE_ERROR when conclusion is absent but outputErrors exist", async () => {
+      const mockFetch = vi.fn().mockResolvedValue({ ok: true, status: 200, statusText: "OK" });
+      vi.stubGlobal("fetch", mockFetch);
+
+      process.env.GH_AW_OTLP_ENDPOINTS = JSON.stringify([{ url: "https://traces.example.com" }]);
+      // GH_AW_AGENT_CONCLUSION intentionally not set (simulates agent job post-step)
+
+      readFileSpy.mockImplementation(filePath => {
+        if (filePath === "/tmp/gh-aw/agent_output.json") {
+          return JSON.stringify({ errors: [{ message: "push_to_pull_request_branch:push failed" }] });
+        }
+        throw Object.assign(new Error("ENOENT"), { code: "ENOENT" });
+      });
+
+      await sendJobConclusionSpan("gh-aw.job.conclusion");
+
+      const body = JSON.parse(mockFetch.mock.calls[0][1].body);
+      const span = body.resourceSpans[0].scopeSpans[0].spans[0];
+      const attrs = Object.fromEntries(span.attributes.map(a => [a.key, a.value.intValue ?? a.value.stringValue ?? a.value.boolValue]));
+      expect(attrs["gh-aw.run.status"]).toBe("failure");
+      expect(span.status.code).toBe(2);
+      expect(span.status.message).toBe("errors detected: push_to_pull_request_branch:push failed");
+    });
+
+    it("includes the first error message in statusMessage when conclusion is absent", async () => {
+      const mockFetch = vi.fn().mockResolvedValue({ ok: true, status: 200, statusText: "OK" });
+      vi.stubGlobal("fetch", mockFetch);
+
+      process.env.GH_AW_OTLP_ENDPOINTS = JSON.stringify([{ url: "https://traces.example.com" }]);
+
+      readFileSpy.mockImplementation(filePath => {
+        if (filePath === "/tmp/gh-aw/agent_output.json") {
+          return JSON.stringify({ errors: [{ message: "Rate limit exceeded" }, { message: "second error" }] });
+        }
+        throw Object.assign(new Error("ENOENT"), { code: "ENOENT" });
+      });
+
+      await sendJobConclusionSpan("gh-aw.job.conclusion");
+
+      const body = JSON.parse(mockFetch.mock.calls[0][1].body);
+      const span = body.resourceSpans[0].scopeSpans[0].spans[0];
+      expect(span.status.message).toBe("errors detected: Rate limit exceeded");
+    });
+
+    it("does not override explicit GH_AW_AGENT_CONCLUSION=success even when outputErrors exist", async () => {
+      const mockFetch = vi.fn().mockResolvedValue({ ok: true, status: 200, statusText: "OK" });
+      vi.stubGlobal("fetch", mockFetch);
+
+      process.env.GH_AW_OTLP_ENDPOINTS = JSON.stringify([{ url: "https://traces.example.com" }]);
+      process.env.GH_AW_AGENT_CONCLUSION = "success";
+
+      readFileSpy.mockImplementation(filePath => {
+        if (filePath === "/tmp/gh-aw/agent_output.json") {
+          return JSON.stringify({ errors: [{ message: "non-fatal warning error" }] });
+        }
+        throw Object.assign(new Error("ENOENT"), { code: "ENOENT" });
+      });
+
+      await sendJobConclusionSpan("gh-aw.job.conclusion");
+
+      const body = JSON.parse(mockFetch.mock.calls[0][1].body);
+      const span = body.resourceSpans[0].scopeSpans[0].spans[0];
+      const attrs = Object.fromEntries(span.attributes.map(a => [a.key, a.value.intValue ?? a.value.stringValue ?? a.value.boolValue]));
+      expect(attrs["gh-aw.run.status"]).toBe("success");
+      expect(span.status.code).toBe(1);
+    });
+
+    it("keeps gh-aw.run.status=success and STATUS_CODE_OK when conclusion is absent and there are no outputErrors", async () => {
+      const mockFetch = vi.fn().mockResolvedValue({ ok: true, status: 200, statusText: "OK" });
+      vi.stubGlobal("fetch", mockFetch);
+
+      process.env.GH_AW_OTLP_ENDPOINTS = JSON.stringify([{ url: "https://traces.example.com" }]);
+      // No GH_AW_AGENT_CONCLUSION and no errors
+
+      readFileSpy.mockImplementation(filePath => {
+        if (filePath === "/tmp/gh-aw/agent_output.json") {
+          return JSON.stringify({ items: [{ type: "pull_request" }] });
+        }
+        throw Object.assign(new Error("ENOENT"), { code: "ENOENT" });
+      });
+
+      await sendJobConclusionSpan("gh-aw.job.conclusion");
+
+      const body = JSON.parse(mockFetch.mock.calls[0][1].body);
+      const span = body.resourceSpans[0].scopeSpans[0].spans[0];
+      const attrs = Object.fromEntries(span.attributes.map(a => [a.key, a.value.intValue ?? a.value.stringValue ?? a.value.boolValue]));
+      expect(attrs["gh-aw.run.status"]).toBe("success");
+      expect(span.status.code).toBe(1);
+    });
+  });
+
   describe("rate-limit enrichment in conclusion span", () => {
     let readFileSpy;