[FEAT] Support clone of cloud related entities - agentic PS, lookups and manual review

README.md

@@ -150,6 +150,17 @@ failed run can be resumed by re-running the same command.
 | `1` | Completed with failures — see the printed report. |
 | `2` | Could not run (setup error or `--on-name-conflict=abort` collision). |
 
+#### Compatibility
+
+`unstract clone` is capability-probed: each phase checks for its endpoint on the
+source and target, and clones only what both orgs support. A capability missing on
+either side is reported and skipped — the run never fails because of a version
+difference. Cloning a newer source into an older target therefore drops the entity
+types the target lacks (listed in the end-of-run report).
+
+- Run the source and target on the same (or a newer-target) Unstract build.
+- Use `unstract-client >= 1.4.0`, the first release that ships `unstract clone`.
+
 ## Questions and Feedback
 
 On Slack, [join great conversations](https://join-slack.unstract.com/) around LLMs, their ecosystem and leveraging them to automate the previously unautomatable!

src/unstract/clone/context.py

@@ -145,3 +145,20 @@ class CloneContext:
     # touches them once per endpoint, never per resource).
     share_cache: dict[str, Any] = field(default_factory=dict)
     share_cache_lock: threading.Lock = field(default_factory=threading.Lock)
+    # Capability-probe memo: (id(client), feature_path) -> present?. Probed
+    # once per (deployment, feature) so cloud-phase gating costs one GET total.
+    probe_cache: dict[tuple[int, str], bool] = field(default_factory=dict)
+
+    def feature_present(self, client: "PlatformClient", path: str) -> bool:
+        """Is ``path`` (a feature's list endpoint) installed on ``client``?
+
+        Memoised per run. Plain dict, no lock — probing runs in the
+        single-threaded orchestrator loop, before any parallel_map fan-out.
+        """
+        key = (id(client), path)
+        cached = self.probe_cache.get(key)
+        if cached is not None:
+            return cached
+        present = client.probe(path)
+        self.probe_cache[key] = present
+        return present

src/unstract/clone/orchestrator.py

@@ -19,11 +19,14 @@
 from unstract.clone.exceptions import CloneError
 from unstract.clone.phases import (
     AdapterPhase,
+    AgenticStudioPhase,
     APIDeploymentPhase,
     ConnectorPhase,
     CustomToolPhase,
     FilesPhase,
     GroupPhase,
+    LookupsPhase,
+    ManualReviewPhase,
     PipelinePhase,
     TagPhase,
     ToolInstancePhase,
@@ -46,18 +49,58 @@
 PHASES: list[tuple[str, type[Phase]]] = [
     ("group", GroupPhase),
     ("adapter", AdapterPhase),
+    # Cloud-only; standalone (own project + registry) and FKs four adapters.
+    # Probe-gated: auto-skips on OSS deployments via ``probe_path``.
+    ("agentic_studio", AgenticStudioPhase),
     ("connector", ConnectorPhase),
     ("tag", TagPhase),
     ("custom_tool", CustomToolPhase),
     ("files", FilesPhase),
+    # Cloud-only; consumes custom_tool's prompt + adapter remaps. Probe-gated:
+    # auto-skips on OSS deployments via ``probe_path``.
+    ("lookups", LookupsPhase),
     ("workflow", WorkflowPhase),
+    # Cloud-only; review rules and settings bind to the workflow.
+    # Probe-gated: auto-skips on OSS deployments via ``probe_path``.
+    ("manual_review", ManualReviewPhase),
     ("tool_instance", ToolInstancePhase),
     ("workflow_endpoint", WorkflowEndpointPhase),
     ("pipeline", PipelinePhase),
     ("api_deployment", APIDeploymentPhase),
 ]
 
 
+def _cloud_phase_runnable(
+    ctx: CloneContext, report: CloneReport, name: str, probe_path: str
+) -> bool:
+    """Decide whether a cloud-only phase should run on this deployment pair.
+
+    Probe source first; only probe target if source has the feature. A probe
+    failure (unexpected status / transport) must not abort an otherwise-fine
+    run — treat it like target-absent: warn + skip, never raise.
+    """
+    try:
+        if not ctx.feature_present(ctx.source, probe_path):
+            # OSS source: behave exactly as if this phase didn't exist.
+            logger.debug("Phase '%s' skipped: feature absent on source", name)
+            return False
+        target_present = ctx.feature_present(ctx.target, probe_path)
+    except Exception as e:
+        msg = f"Phase '{name}' skipped: capability probe failed ({e})"
+        logger.warning(msg)
+        report.warnings.append(msg)
+        return False
+    if not target_present:
+        msg = (
+            f"Phase '{name}' skipped: feature present on source but not on "
+            "target deployment"
+        )
+        logger.warning(msg)
+        report.warnings.append(msg)
+        return False
+    return True
+
+
 def clone(
     source: OrgEndpoint,
     target: OrgEndpoint,
@@ -93,6 +136,11 @@ def clone(
                 report.skipped_phases.append(name)
                 logger.info("Phase '%s' skipped (excluded)", name)
                 continue
+            probe_path = getattr(phase_cls, "probe_path", None)
+            if probe_path is not None and not _cloud_phase_runnable(
+                ctx, report, name, probe_path
+            ):
+                continue
             logger.info("=== Phase: %s ===", name)
             phase_started = time.perf_counter()
             try:

src/unstract/clone/phases/__init__.py

@@ -8,12 +8,15 @@
 """
 
 from unstract.clone.phases.adapter import AdapterPhase
+from unstract.clone.phases.agentic_studio import AgenticStudioPhase
 from unstract.clone.phases.api_deployment import APIDeploymentPhase
 from unstract.clone.phases.base import Phase
 from unstract.clone.phases.connector import ConnectorPhase
 from unstract.clone.phases.custom_tool import CustomToolPhase
 from unstract.clone.phases.files import FilesPhase
 from unstract.clone.phases.group import GroupPhase
+from unstract.clone.phases.lookups import LookupsPhase
+from unstract.clone.phases.manual_review import ManualReviewPhase
 from unstract.clone.phases.pipeline import PipelinePhase
 from unstract.clone.phases.tag import TagPhase
 from unstract.clone.phases.tool_instance import ToolInstancePhase
@@ -23,10 +26,13 @@
 __all__ = [
     "APIDeploymentPhase",
     "AdapterPhase",
+    "AgenticStudioPhase",
     "ConnectorPhase",
     "CustomToolPhase",
     "FilesPhase",
     "GroupPhase",
+    "LookupsPhase",
+    "ManualReviewPhase",
     "Phase",
     "PipelinePhase",
     "TagPhase",

src/unstract/clone/phases/adapter.py

@@ -4,9 +4,8 @@
 against target, POST create if missing, record source->target UUID in the
 remap table for downstream phases.
 
-Frictionless onboarding adapters are excluded — the backend's
-service-account queryset already filters them out, so clone never
-sees them.
+Frictionless onboarding adapters are excluded — they're already filtered
+out of what this org's Platform key can list, so clone never sees them.
 """
 
 from __future__ import annotations