Fix: AgentRuntime tool failure handling and PlannerService fallback graph persistence

AgentRuntime.ExecutePlannerTaskAsync:
- Add PlannerValidationException and FileNotFoundException to the tool
  execution catch clause so that 'Completed tasks are immutable' and
  'Approved path does not exist' errors are returned as tool results
  instead of propagating as unhandled task failures.
- Treat task_update as a metadata tool: its failures no longer add to
  unresolvedToolFailures, so the LLM is not forced into an infinite
  retry loop when task_update receives invalid arguments.
- Skip throwing InvalidOperationException for RetryAllowed=false
  failures originating from metadata tools (task_update), preventing
  unnecessary replan cycles caused by graph bookkeeping errors.
- Improve the 'No tasks were executed' fallback message: when replan
  was attempted but no tasks could execute, report that the sandbox
  policy denied required operations.

PlannerService.PlanAsync:
- Persist fallback TaskGraph to session.TaskGraphs[taskId] in all
  three fallback paths (ChatStructuredAsync exception, empty/null plan,
  BuildTaskGraph exception) so that BeginTask's TryAdd does not create
  an empty graph that shadows the fallback.

1 files changed, 9 insertions, 4 deletions

diff --git a/src/Agent.Core/Runtime/AgentRuntime.cs b/src/Agent.Core/Runtime/AgentRuntime.cs
index aa81486..0b261da 100644
--- a/src/Agent.Core/Runtime/AgentRuntime.cs
+++ b/src/Agent.Core/Runtime/AgentRuntime.cs
@@ -110,9 +110,12 @@ public sealed class AgentRuntime(TaskRegistry registry, IOllamaClient ollama, Ag
             }
 
             // All tasks completed — assemble final result.
+            // If replan was attempted but no tasks could execute, provide a meaningful message.
             string finalContent = finalContents.Count > 0
                 ? string.Join("\n\n", finalContents)
-                : "No tasks were executed.";
+                : replanCount > 0
+                    ? "The request could not be completed because the sandbox policy denied required operations."
+                    : "No tasks were executed.";
             string? cleanupFailure = await EndTaskSafelyAsync(taskId, true, CancellationToken.None).ConfigureAwait(false);
             if (cleanupFailure is not null) throw new InvalidOperationException(cleanupFailure);
             session.Append(new ConversationMessage("user", task.Request));
@@ -210,20 +213,22 @@ public sealed class AgentRuntime(TaskRegistry registry, IOllamaClient ollama, Ag
                     {
                         toolResult = new(false, ex.Message, false, $"{call.Function.Name}:{ex.GetType().Name}:{ex.Message}", RetryAllowed: false);
                     }
-                    catch (Exception ex) when (ex is ToolValidationException or ArgumentException)
+                    catch (Exception ex) when (ex is ToolValidationException or ArgumentException or PlannerValidationException or FileNotFoundException)
                     {
                         toolResult = new(false, ex.Message, false, $"{call.Function.Name}:{ex.GetType().Name}:{ex.Message}");
                     }
 
                     externalEffectEvidence |= toolResult.Success && toolResult.ExternalEffect;
+                    // task_update is a metadata tool for graph bookkeeping; its failures should not block completion.
+                    bool isMetadataTool = call.Function.Name == "task_update";
                     if (toolResult.Success) unresolvedToolFailures.Remove(call.Function.Name);
-                    else unresolvedToolFailures[call.Function.Name] = toolResult.FailureFingerprint ?? toolResult.Output;
+                    else if (!isMetadataTool) unresolvedToolFailures[call.Function.Name] = toolResult.FailureFingerprint ?? toolResult.Output;
 
                     // Record ExecutionResult in session for context summary.
                     session.AppendExecutionResult(new ExecutionResult(taskId, call.Function.Name, toolResult.Success, toolResult.Output));
 
                     _observer.ToolResult(taskId, call.Function.Name, toolResult);
-                    if (!toolResult.Success && !toolResult.RetryAllowed)
+                    if (!toolResult.Success && !toolResult.RetryAllowed && !isMetadataTool)
                         throw new InvalidOperationException($"Tool '{call.Function.Name}' was denied by sandbox policy and will not be retried: {toolResult.Output}");
                     failureFingerprint = toolResult.FailureFingerprint ?? failureFingerprint;
                     messages.Add(new OllamaMessage("tool", toolResult.Output, ToolName: call.Function.Name));