fix(clone): address PR #15 P1s — workflow detail GET + remap snapshot

- WorkflowPhase now does list → per-id GET → POST like other phases,
  so source_settings/destination_settings JSON blobs (carrying
  connector UUIDs) aren't dropped by stripped list serializers.
- RemapTable.resolve_any iterates over a snapshot to avoid
  `dictionary changed size during iteration` against a concurrent
  record() from sibling workers.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Author: chandrasekharan-zipstack

src/unstract/clone/context.py

@@ -79,7 +79,9 @@ def resolve(self, entity: str, src_uuid: str) -> str | None:
         return self._table.get(entity, {}).get(src_uuid)
 
     def resolve_any(self, src_uuid: str) -> str | None:
-        for mapping in self._table.values():
+        # Snapshot to avoid `RuntimeError: dictionary changed size during
+        # iteration` when a concurrent record() inserts a new entity bucket.
+        for mapping in list(self._table.values()):
             hit = mapping.get(src_uuid)
             if hit is not None:
                 return hit

src/unstract/clone/phases/workflow.py

@@ -127,7 +127,20 @@ def _clone_one(
             logger.info("[dry-run] would create workflow '%s' src=%s", name, src_id)
             return
         else:
-            remapped = remap_uuids(src, self.ctx.remap)
+            # List endpoints serve stripped payloads (e.g. AdapterListSerializer
+            # omits adapter_metadata_b); workflow detail carries the JSON blobs
+            # source_settings / destination_settings that embed connector UUIDs.
+            try:
+                src_detail = self.ctx.source.get_workflow(src_id)
+            except Exception as e:
+                logger.exception(
+                    "Failed to GET source workflow %s detail: %s", name, e
+                )
+                with lock:
+                    result.failed += 1
+                    result.errors.append(f"GET source detail {name}: {e}")
+                return
+            remapped = remap_uuids(src_detail, self.ctx.remap)
             payload = build_post_payload(remapped, self._writable)
             try:
                 tgt = self.ctx.target.create_workflow(payload)

tests/clone/test_workflow_phase.py

@@ -53,6 +53,12 @@ def list_workflows(self, *, name: str | None = None):
             result = [w for w in result if w["workflow_name"] == name]
         return list(result)
 
+    def get_workflow(self, workflow_id: str) -> dict:
+        for w in self.workflows:
+            if w["id"] == workflow_id:
+                return dict(w)
+        raise KeyError(workflow_id)
+
     def list_tool_instances(self, *, workflow_id: str | None = None) -> list[dict]:
         if workflow_id is None:
             return list(self.tool_instances)