UiPath · apetraru-uipath · Aug 1, 2026 · Aug 1, 2026 · Aug 1, 2026 · Aug 1, 2026
@@ -50,6 +50,10 @@
 
 `Key` identifies the backing app. `resource list` does not return `systemName`, `deployVersion`, or the action schema — fetch those in Step 3 with `uip solution resources get`.
 
+**Solution-internal apps require the UNFILTERED list.** `--kind`, `--source`, and `--search` filters currently apply only to the remote catalog — a `Source: "Local"` app is dropped by every filtered query and appears only in a plain `uip solution resources list --output json`. Run the unfiltered list (from the solution directory) before concluding an app is absent, and always pass the row's `Key` — not the name — to `resources get` (name lookups fail for solution-internal apps).
+
+**App not in the list — STOP.** If the named app has no row in `uip solution resources list` (or Step 3's `resources get` exits non-zero / "was not found"), the app cannot back an escalation. Report `<APP_NAME> was not found in the solution or the resource catalog — the escalation cannot be wired` and stop. Do NOT retry with alternate search terms, other endpoints, raw API calls, or re-authentication, and do NOT write the escalation (or an escalate guardrail action) referencing the missing app — a single failed lookup is terminal for this run.
+
 ### Step 3 — Fetch the app's action schema
 
 Pass the `Key` from Step 2 to `uip solution resources get` — one CLI-native call returns the app spec, including the action schema. Works for both external (`Source: "Remote"`) and solution-internal (`Source: "Local"`) apps, even before the app is imported into the solution. Run from the solution directory (or pass `--solution-folder <path>`).
@@ -170,7 +174,7 @@
      "properties": {
        "resourceKey": "<appId-guid>",              // the `Key` from `uip solution resources list` (== the key passed to `resources get`)
        "appName": "<appName>",                     // `Name` from `uip solution resources list` (== `Spec.Name` from `resources get`)
        "folderName": "Shared/Approvals",           // literal Folder from `uip solution resources list`. External: Orchestrator folder (e.g., "Shared/Approvals"). Solution-internal: "solution_folder". uip agent refresh translates this to folderPath in the App binding inside bindings_v2.json.
        "appVersion": 1,                            // integer — the `version` field of the parsed `Spec.ActionSchema` (NOT `Spec.Version`, the package semver)
        "isActionableMessageEnabled": false,
        "actionableMessageMetaData": null
@@ -229,7 +233,7 @@

 ### Step 6b — Inline agents only: wire the escalation flow node

 **Skip if the agent is standalone.** If the escalation is on an **inline** agent (embedded in a flow), the `resource.json` alone is never reached at runtime — you MUST also add an escalation flow node connected to the autonomous node's `escalation` handle. The registry exposes the escalation node as concrete variants (e.g. `uipath.agent.resource.escalation.coded-action-app` for an Action Center / Workflow Action app) — there is no bare `uipath.agent.resource.escalation` node. Discover the available variant with `uip maestro flow registry search "escalation" --output json` (pick the one with `AvailableOnTenant: true`), fetch its manifest with `uip maestro flow registry get <NodeType> --output json`, then hand the node + edge authoring to the `uipath-maestro-flow` skill (Critical Rule 16 — this skill does not author `.flow` graphs directly). Run Step 7's refresh/validate with `--inline-in-flow` plus `--bindings-target <FlowProjectDir>/bindings_v2.json`. See [../inline-in-flow/inline-in-flow.md](../inline-in-flow/inline-in-flow.md).

 ### Step 7 — Refresh, validate, and refresh solution resources

@@ -256,7 +260,7 @@
 ### Step 8 — Bundle and upload

 ```bash
 uip solution bundle . -d ./dist --output json
 uip solution upload ./dist/<SOLUTION_NAME>.uis --output json
 ```


@@ -171,6 +171,14 @@ cases:
 Name the catalog-recommended action for that scope in the description. Severity `judgment` — a guardrail that breaks
 the agent or leaves a security gap can be Critical; a milder ineffectiveness is Warning/Info.
 
+**The catalog verdict IS the evidence — do not verify SDK behavior empirically.** Source read + catalog comparison
+fully decides this check. Do NOT import or execute the project's dependencies, inspect `uipath_langchain` /
+`langchain` package sources, fetch SDK documentation sites, or run the agent's factory code to "prove" an action is
+ineffective — that forensic detour adds no admissible evidence to the finding and routinely burns the entire review
+budget. Once the configured action diverges from the catalog's recommendation for that scope, write the finding and
+move on to the report (the report file is the deliverable; an unwritten report scores zero regardless of analysis
+depth).
+
 ### Relevance + Wiring Check → `CODED_GUARDRAIL_MISAPPLIED`
 
 Two ways a format-valid guardrail can be misapplied:

@@ -0,0 +1 @@
+{"SolutionId": "e5ca1a7e-0000-4000-8000-000000000000", "Projects": []}
@@ -0,0 +1,7 @@
+{
+  "DocVersion": "1.0.0",
+  "StudioMinVersion": "2025.04.0",
+  "SolutionId": "e5ca1a7e-0000-4000-8000-000000000000",
+  "Projects": [],
+  "AutomationHubIdeaUrl": null
+}
@@ -0,0 +1,45 @@
+{
+  "docVersion": "1.0.0",
+  "resource": {
+    "name": "EscalationWorksApp",
+    "kind": "app",
+    "type": "workflow Action",
+    "apiVersion": "apps.uipath.com/v1",
+    "projectKey": "e5ca1a7e-0005-4005-8005-000000000005",
+    "isOverridable": true,
+    "dependencies": [
+      {
+        "name": "EscalationWorksApp",
+        "kind": "appVersion"
+      },
+      {
+        "name": "EscalationWorksApp",
+        "kind": "process"
+      }
+    ],
+    "runtimeDependencies": [],
+    "files": [],
+    "folders": [
+      {
+        "fullyQualifiedName": "solution_folder"
+      }
+    ],
+    "spec": {
+      "name": "EscalationWorksApp",
+      "description": "Escalation works queue app (generic single input/output form).",
+      "type": "workflow Action",
+      "version": null,
+      "appSystemName": null,
+      "appVersionRef": {
+        "key": "e5ca1a7e-0002-4002-8002-000000000002"
+      },
+      "codeBehindProcess": {
+        "key": "e5ca1a7e-0003-4003-8003-000000000003"
+      },
+      "actionSchema": "{\"version\": 1, \"inputs\": [{\"name\": \"input\", \"type\": \"System.String\", \"description\": \"Generic input value\"}], \"inOuts\": [], \"outputs\": [{\"name\": \"output\", \"type\": \"System.String\", \"description\": \"Generic output value\"}], \"outcomes\": [{\"name\": \"Submit\"}, {\"name\": \"Reject\"}]}",
+      "manualActivationActions": null
+    },
+    "locks": [],
+    "key": "e5ca1a7e-0001-4001-8001-000000000001"
+  }
+}
@@ -0,0 +1,30 @@
+{
+  "docVersion": "1.0.0",
+  "resource": {
+    "name": "EscalationWorksApp",
+    "kind": "appVersion",
+    "apiVersion": "apps.uipath.com/v2",
+    "projectKey": "e5ca1a7e-0005-4005-8005-000000000005",
+    "isOverridable": true,
+    "dependencies": [],
+    "runtimeDependencies": [],
+    "files": [],
+    "folders": [
+      {
+        "fullyQualifiedName": "solution_folder"
+      }
+    ],
+    "spec": {
+      "name": "EscalationWorksApp",
+      "description": null,
+      "publishNote": null,
+      "version": null,
+      "projectId": null,
+      "isAppPublic": false,
+      "packageFile": null,
+      "packageKey": null
+    },
+    "locks": [],
+    "key": "e5ca1a7e-0002-4002-8002-000000000002"
+  }
+}
@@ -0,0 +1,26 @@
+{
+  "docVersion": "1.0.0",
+  "resource": {
+    "name": "EscalationWorksApp",
+    "kind": "package",
+    "apiVersion": "orchestrator.uipath.com/v1",
+    "projectKey": "e5ca1a7e-0005-4005-8005-000000000005",
+    "isOverridable": true,
+    "dependencies": [],
+    "runtimeDependencies": [],
+    "files": [],
+    "folders": [
+      {
+        "fullyQualifiedName": "solution_folder"
+      }
+    ],
+    "spec": {
+      "fileName": null,
+      "fileReference": null,
+      "name": "EscalationWorksApp",
+      "description": null
+    },
+    "locks": [],
+    "key": "e5ca1a7e-0004-4004-8004-000000000004"
+  }
+}
@@ -0,0 +1,63 @@
+{
+  "docVersion": "1.0.0",
+  "resource": {
+    "name": "EscalationWorksApp",
+    "kind": "process",
+    "type": "webApp",
+    "apiVersion": "orchestrator.uipath.com/v1",
+    "projectKey": "e5ca1a7e-0005-4005-8005-000000000005",
+    "isOverridable": true,
+    "dependencies": [
+      {
+        "name": "EscalationWorksApp",
+        "kind": "package"
+      }
+    ],
+    "runtimeDependencies": [],
+    "files": [],
+    "folders": [
+      {
+        "fullyQualifiedName": "solution_folder"
+      }
+    ],
+    "spec": {
+      "entryPointUniqueId": null,
+      "inputArgumentsSchema": null,
+      "outputArgumentsSchema": null,
+      "inputArgumentsSchemaV2": null,
+      "outputArgumentsSchemaV2": null,
+      "type": "WebApp",
+      "name": "EscalationWorksApp",
+      "description": null,
+      "package": {
+        "key": "e5ca1a7e-0004-4004-8004-000000000004"
+      },
+      "packageName": "ValidationSol.webApp.EscalationWorksApp",
+      "packageVersion": null,
+      "entryPointName": null,
+      "inputArguments": null,
+      "jobPriority": "Medium",
+      "hiddenForAttendedUser": true,
+      "alwaysRunning": true,
+      "autoStartProcess": false,
+      "jobRecording": "Disabled",
+      "targetFrameworkValue": "Portable",
+      "duration": 40,
+      "frequency": 500,
+      "quality": 100,
+      "remoteControlAccess": "None",
+      "robotSize": "Automatic",
+      "retentionAction": "Delete",
+      "retentionPeriod": 30,
+      "retentionBucketRef": null,
+      "staleRetentionAction": "Delete",
+      "staleRetentionPeriod": 180,
+      "staleRetentionBucketRef": null,
+      "entryPoints": null,
+      "connections": null,
+      "tags": []
+    },
+    "locks": [],
+    "key": "e5ca1a7e-0003-4003-8003-000000000003"
+  }
+}
@@ -1,32 +1,42 @@
 task_id: skill-agent-guardrail-escalation-app-validation
 description: >
-  Guardrail escalation app action schema validation. Asks the agent to add
-  an escalate guardrail using an incompatible Action Center app
-  ("EscalationWorksApp") that does NOT have the required guardrail schema
-  (missing 8 required inputs, 3 required outputs, wrong outcome names).
-  The agent must validate the action schema and reject the app with the
-  error message instead of writing the guardrail.
+  Guardrail escalation app action schema validation. The fixture ships a
+  "ValidationSol" solution containing a solution-internal Action Center app
+  ("EscalationWorksApp") whose action schema is deliberately incompatible
+  with the guardrail escalation contract (generic input/output instead of
+  the 8 required inputs and 3 required outputs; Submit/Reject instead of
+  Approve/Reject outcomes). The agent is asked to add an escalate guardrail
+  backed by that app; it must fetch the action schema via
+  `uip solution resources get`, detect the mismatch, and reject the app with
+  the error message instead of writing the guardrail.
 tags: [uipath-agents, e2e, mode:build, lifecycle:generate, low-code, guardrail]
+
+pre_run:
+  - command: "cp -r $SKILLS_REPO_PATH/tests/tasks/uipath-agents/lowcode/guardrails/escalation_app_validation/_fixtures/ValidationSol ."
+    timeout: 10
+
 initial_prompt: |
   Create a UiPath low-code agent "ReviewAgent" that handles compliance review
-  requests. Add a guardrail that watches for personal data (email addresses)
-  and, instead of blocking, escalates flagged cases to a human reviewer using
-  our "EscalationWorksApp" Action Center app. The solution should be named
-  "ValidationSol".
+  requests, inside the existing "ValidationSol" solution in the current
+  directory (do NOT re-create the solution). Add a guardrail that watches for
+  personal data (email addresses) and, instead of blocking, escalates flagged
+  cases to a human reviewer using our "EscalationWorksApp" Action Center app,
+  which is already part of this solution.
 
-  Do NOT publish, upload, or deploy. Do NOT ask for approval,
-  confirmation, or feedback. Do NOT pause between planning and
-  implementation. Do NOT write a plan file. Complete the entire task
-  end-to-end in a single pass.
+  Do NOT publish, upload, or deploy. Do NOT run
+  `uip solution resources refresh` — the solution resources are already in
+  sync. Do NOT ask for approval, confirmation, or feedback. Do NOT pause
+  between planning and implementation. Do NOT write a plan file. Complete
+  the entire task end-to-end in a single pass.
 
 success_criteria:
-  - type: command_executed
-    description: "Agent created the solution"
+  - type: command_not_executed
+    description: "Agent did NOT re-initialize the pre-existing ValidationSol solution"
     tool_name: "Bash"
     command_pattern: 'uip\s+solution\s+init'
-    min_count: 1
+    exclude_pattern: '--help'
     weight: 1.0
-    pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
+    pass_threshold: 1.0
 
   - type: command_executed
     description: "Agent scaffolded the agent project"
@@ -62,5 +72,6 @@ success_criteria:
 
 run_limits:
   expected_turns: 20
-  max_turns: 40
-  turn_timeout: 1200
+  max_turns: 60
+  turn_timeout: 1800
+  task_timeout: 2400
@@ -34,6 +34,7 @@
 """
 
 import json
+import re
 from pathlib import Path
 
 SOLUTION_ID = "11111111-1111-4111-8111-111111111111"
@@ -124,6 +125,98 @@ def _write_json(path: Path, data: dict) -> None:
     path.write_text(json.dumps(data, indent=2), encoding="utf-8")
 
 
+_TOKEN_RE = re.compile(r"\{\{\s*([^}]+?)\s*\}\}")
+
+
+def content_tokens(content: str) -> list[dict]:
+    """Build the ``contentTokens`` array matching a message ``content`` string.
+
+    `uip agent validate` requires contentTokens to mirror content segment by
+    segment: plain text becomes ``simpleText`` entries and each ``{{expr}}``
+    becomes a ``variable`` entry. Inject scripts that patch a message's
+    ``content`` MUST also set ``contentTokens = content_tokens(content)`` or
+    the fixture fails schema validation before the reviewer ever sees the
+    intended defect signal.
+    """
+    tokens: list[dict] = []
+    pos = 0
+    for m in _TOKEN_RE.finditer(content):
+        if m.start() > pos:
+            tokens.append({"type": "simpleText", "rawString": content[pos : m.start()]})
+        tokens.append({"type": "variable", "rawString": m.group(1)})
+        pos = m.end()
+    if pos < len(content):
+        tokens.append({"type": "simpleText", "rawString": content[pos:]})
+    return tokens
+
+
+def set_message(agent_data: dict, role: str, content: str) -> None:
+    """Set a message's content AND its matching contentTokens in one step."""
+    for msg in agent_data.get("messages", []):
+        if msg.get("role") == role:
+            msg["content"] = content
+            msg["contentTokens"] = content_tokens(content)
+
+
+def connection_binding(connection_id: str, connector_key: str) -> dict:
+    """One bindings_v2 entry for an integration tool's connection.
+
+    Mirrors the packager's generateBindings() output — `uip agent validate`
+    derives the expected entries from the tool resources and fails with
+    "expected binding entry ... is not present on disk" when they're absent.
+    """
+    return {
+        "resource": "connection",
+        "key": connection_id,
+        "value": {
+            "connectionId": {
+                "defaultValue": connection_id,
+                "isExpression": False,
+                "displayName": "Connection ID",
+            }
+        },
+        "metadata": {
+            "connector": connector_key,
+            "useConnectionService": "true",
+            "bindingsVersion": "2.2",
+            "solutionsSupport": "true",
+        },
+    }
+
+
+def process_binding(process_name: str, folder_path: str = "Shared") -> dict:
+    """One bindings_v2 entry for a process tool (see connection_binding)."""
+    return {
+        "resource": "process",
+        "key": process_name,
+        "value": {
+            "name": {
+                "defaultValue": process_name,
+                "isExpression": False,
+                "displayName": "Process name",
+            },
+            "folderPath": {
+                "defaultValue": folder_path,
+                "isExpression": False,
+                "displayName": "Folder path",
+            },
+        },
+        "metadata": {
+            "subType": "process",
+            "bindingsVersion": "2.2",
+            "solutionsSupport": "true",
+        },
+    }
+
+
+def write_bindings(project_dir: Path, entries: list[dict]) -> None:
+    """Write bindings_v2.json (root) + .agent-builder/bindings.json with the
+    given entries, matching what `uip agent refresh` would generate."""
+    data = {"version": "2.0", "resources": entries}
+    _write_json(Path(project_dir) / "bindings_v2.json", data)
+    _write_json(Path(project_dir) / ".agent-builder" / "bindings.json", data)
+
+
 def write_baseline_lowcode_agent(
     solution_root: Path,
     *,
@@ -176,6 +269,9 @@ def write_baseline_lowcode_agent(
         },
     )
     _write_json(project_dir / "flow-layout.json", {"zoom": 1.0})
+    # Root-level bindings_v2.json — real `uip agent init` projects carry it,
+    # and `uip agent validate` requires it as soon as resources/ exists.
+    _write_json(project_dir / "bindings_v2.json", {"version": "2.0", "resources": []})
 
     # --- .agent-builder/ regenerated artifacts (same logical content)
     agent_builder = project_dir / ".agent-builder"
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		{"SolutionId": "e5ca1a7e-0000-4000-8000-000000000000", "Projects": []}