Add checkpoint persistence to agent core loop

coreLoop now saves incremental checkpoints after each tool-call turn
and checks a cooperative stop signal at turn boundaries. SuspendedError
is handled in finishRun, executeParallel, and executeSingleTool.
Approval-interrupted checkpoints are persisted for both flat and
nested interruptions.

Introduce RunOption, WithCheckpointStore, RunWithOpts, ResumeWithOpts,
and RunStreamedWithOpts so callers can provide checkpoint storage.
Add StreamEventSuspended and OnRunRestore hook.

Signed-off-by: Aurélien Sibiril <81782+aureliensibiril@users.noreply.github.com>

pkg/agent/stream.go

@@ -49,6 +49,7 @@ const (
 	StreamEventToolEnd    StreamEventType = "tool_end"
 	StreamEventHandoff    StreamEventType = "handoff"
 	StreamEventComplete   StreamEventType = "complete"
+	StreamEventSuspended  StreamEventType = "suspended"
 	StreamEventError      StreamEventType = "error"
 )
 
@@ -59,6 +60,10 @@ func (sr *StreamedRun) Wait() (*Result, error) {
 }
 
 func (a *Agent) RunStreamed(ctx context.Context, messages []llm.Message) *StreamedRun {
+	return a.RunStreamedWithOpts(ctx, messages)
+}
+
+func (a *Agent) RunStreamedWithOpts(ctx context.Context, messages []llm.Message, opts ...RunOption) *StreamedRun {
 	events := make(chan StreamEvent, 64)
 	sr := &StreamedRun{
 		Events: events,
@@ -69,17 +74,17 @@ func (a *Agent) RunStreamed(ctx context.Context, messages []llm.Message) *Stream
 		defer close(sr.done)
 		defer close(events)
 
-		result, err := coreLoop(
-			ctx,
-			a,
-			messages,
-			runOpts{
-				callLLM: streamingCallLLM(events),
-				onEvent: func(ctx context.Context, ev StreamEvent) {
-					trySendEvent(ctx, events, ev)
-				},
+		ro := runOpts{
+			callLLM: streamingCallLLM(events),
+			onEvent: func(ctx context.Context, ev StreamEvent) {
+				trySendEvent(ctx, events, ev)
 			},
-		)
+		}
+		for _, opt := range opts {
+			opt(&ro)
+		}
+
+		result, err := coreLoop(ctx, a, messages, ro)
 
 		sr.result = result
 		sr.err = err