feat(team): preserve node run evidence

app-instance/backend/beaver/coordinator/execution/scheduler.py

@@ -241,7 +241,7 @@ class TeamGraphScheduler:
         failed = [item for item in results if not item.success]
         if failed:
             failure_lines = [
-                f"- {item.node_id}: {item.error or item.finish_reason}"
+                f"- {item.node_id}: {item.error or item.finish_reason} evidence={'yes' if item.evidence else 'no'}"
                 for item in failed
             ]
             summary_parts.append("Failed nodes:\n" + "\n".join(failure_lines))

app-instance/backend/beaver/coordinator/local.py

@@ -6,6 +6,7 @@ from uuid import uuid4
 
 from beaver.engine import AgentLoop
 from beaver.engine.providers import ProviderBundle
+from beaver.tasks.evidence import EvidenceBuilder
 
 from .models import DelegationEnvelope, NodeRunResult
 
@@ -47,6 +48,13 @@ class LocalAgentRunner:
             pinned_skill_contexts=envelope.inherited_pinned_skill_contexts,
             allow_candidate_generation=allow_candidate_generation,
         )
+        loaded = self.loop.boot()
+        evidence = EvidenceBuilder(loaded.session_manager).build_run_evidence(
+            result.session_id,
+            result.run_id,
+            result.output_text,
+            result.finish_reason,
+        )
         success = result.finish_reason == "stop"
         return NodeRunResult(
             node_id=envelope.node_id or envelope.agent.name,
@@ -56,6 +64,7 @@ class LocalAgentRunner:
             session_id=result.session_id,
             finish_reason=result.finish_reason,
             error=None if success else (result.output_text or result.finish_reason),
+            evidence=evidence,
         )
 
     @staticmethod

app-instance/backend/beaver/coordinator/models.py

@@ -7,6 +7,7 @@ from typing import TYPE_CHECKING, Any, Literal
 
 if TYPE_CHECKING:
     from beaver.engine.context import SkillContext
+    from beaver.tasks.evidence import RunEvidence
 
 
 TeamStrategy = Literal[
@@ -116,6 +117,7 @@ class NodeRunResult:
     session_id: str | None = None
     finish_reason: str = "stop"
     error: str | None = None
+    evidence: "RunEvidence | None" = None
 
     def to_dict(self) -> dict[str, Any]:
         return {
@@ -126,6 +128,7 @@ class NodeRunResult:
             "session_id": self.session_id,
             "finish_reason": self.finish_reason,
             "error": self.error,
+            "evidence": self.evidence.to_dict() if self.evidence is not None else None,
         }