fix: Isolate tracking failures and log failed judge evaluations in agent

Mirror the managed_model.py fix in managed_agent.py: wrap
tracker.track_judge_result() in try/except so a tracking failure
does not destroy successfully computed evaluation results, and log
a warning when a judge evaluation fails (r.success is False) so
failures are visible rather than silently skipped.

Author: jsonbailey

packages/sdk/server-ai/src/ldai/managed_agent.py

@@ -3,6 +3,7 @@
 import asyncio
 from typing import List
 
+from ldai import log
 from ldai.models import AIAgentConfig
 from ldai.providers.runner import Runner
 from ldai.providers.types import JudgeResult, ManagedResult
@@ -65,7 +66,12 @@ async def _run_and_track(eval_task: asyncio.Task) -> List[JudgeResult]:
             results = await eval_task
             for r in results:
                 if r.success:
-                    tracker.track_judge_result(r)
+                    try:
+                        tracker.track_judge_result(r)
+                    except Exception:
+                        pass
+                else:
+                    log.warning("Judge evaluation failed: %s", r.error_message)
             return results
 
         return asyncio.create_task(_run_and_track(evaluator_task))