fix: a run started a moment ago is never judged dead — another dashboard's tmux list may predate it

Only the process that starts a pane writes it into its own cached tmux list
(noteSession); a second dashboard on the same database (a dev server) holds
a member list up to 15 s old, so its heal beat ended a Codex pane seconds
after it started and the proxy gateway answered 401. reconcileOrphans and
healRunId skip a run younger than 20 s (dashboard-runs.isYoungRun).
Tests: dashboard-runs (young vs old).

server/__tests__/s3-unseen-member-runs.test.js

@@ -31,7 +31,15 @@ const { healAllRunIds } = require("../routes/lanes");
 after(() => fs.rmSync(ROOT, { recursive: true, force: true }));
 
 const start = (id, laneId, cwd) =>
-  runs.recordRun({ id, mode: "interactive", cwd, status: "running", laneId, sessionId: `s-${id}` });
+  runs.recordRun({
+    id,
+    mode: "interactive",
+    cwd,
+    status: "running",
+    laneId,
+    sessionId: `s-${id}`,
+    startedAt: Date.parse("2026-01-01T00:00:00Z"),
+  });
 
 describe("a server without S3 on an S3 machine's database", () => {
   it("never ends a member's run, and still ends its own", () => {