Round MCP timeout durations in error messages (#31612)

## Summary

MCP operation timeout errors currently print the full debug precision of
the remaining timeout budget. That makes a configured 30-second timeout
show up as something like `29.999999875s`.

This PR rounds the displayed duration to a whole unit, so the error is
clean and stable:

```text
timed out awaiting tools/list after 30s
```

The timeout behavior itself is unchanged; this only affects the
human-facing error text. A regression test covers the reported
`tools/list` case.

codex-rs/rmcp-client/src/rmcp_client.rs

@@ -228,7 +228,7 @@ where
 enum ClientOperationError {
     #[error(transparent)]
     Service(#[from] rmcp::service::ServiceError),
-    #[error("timed out awaiting {label} after {duration:?}")]
+    #[error("timed out awaiting {label} after {duration:.0?}")]
     Timeout { label: String, duration: Duration },
 }
 
@@ -1219,6 +1219,16 @@ mod tests {
 
     use super::*;
 
+    #[test]
+    fn client_operation_timeout_rounds_duration() {
+        let error = ClientOperationError::Timeout {
+            label: "tools/list".to_string(),
+            duration: Duration::from_nanos(29_999_999_875),
+        };
+
+        assert_eq!(error.to_string(), "timed out awaiting tools/list after 30s");
+    }
+
     #[tokio::test]
     async fn active_time_timeout_pauses_while_elicitation_is_pending() {
         let pause_state = ElicitationPauseState::new();