From deb3ad0386fd6bf0254dcd0e2928995fff297a12 Mon Sep 17 00:00:00 2001
From: Andrei Petraru <andrei.petraru@uipath.com>
Date: Thu, 9 Jul 2026 16:52:55 +0300
Subject: [PATCH] docs: document LLM-as-judge guardrail validator [AL-469]

Add an "LLM-as-judge" section to the core guardrails docs: a decorator example
plus the guardrail_text/model/threshold/examples parameters and their limits.
Include it in the Execution Stages table (PRE_AND_POST) and add an API Reference
mkdocstrings block for LLMAsJudgeValidator.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---
 packages/uipath/docs/core/guardrails.md | 41 ++++++++++++++++++++++++-
 1 file changed, 40 insertions(+), 1 deletion(-)

diff --git a/packages/uipath/docs/core/guardrails.md b/packages/uipath/docs/core/guardrails.md
index 4c380c20c..6a220bfe2 100644
--- a/packages/uipath/docs/core/guardrails.md
+++ b/packages/uipath/docs/core/guardrails.md
@@ -95,7 +95,7 @@ The `stage` parameter controls when the guardrail evaluates. Not all validators
 |-------|---------------|--------------|
 | `PRE` | Before the function runs | All validators |
 | `POST` | After the function runs | All except `UserPromptAttacksValidator` |
-| `PRE_AND_POST` | Both before and after | `PIIValidator`, `HarmfulContentValidator`, `CustomValidator` |
+| `PRE_AND_POST` | Both before and after | `PIIValidator`, `HarmfulContentValidator`, `LLMAsJudgeValidator`, `CustomValidator` |
 
 ## Built-in Validators
 
@@ -208,6 +208,40 @@ def generate_code(spec: str) -> str:
     ...
 ```
 
+### LLM-as-judge
+
+Evaluates content against a rule written in plain language, using a judge LLM to decide whether the payload complies. Use it for policy checks that are hard to express as fixed entities or rules — tone, topicality, disclaimers, domain-specific policies.
+
+```python
+from uipath.platform.guardrails import (
+    BlockAction,
+    GuardrailExecutionStage,
+    LLMAsJudgeValidator,
+    guardrail,
+)
+
+@guardrail(
+    validator=LLMAsJudgeValidator(
+        guardrail_text=(
+            "The response must remain professional and must not contain "
+            "financial or investment advice."
+        ),
+        model="gpt-4o-2024-08-06",
+        threshold=2,
+    ),
+    action=BlockAction(),
+    name="No financial advice",
+    stage=GuardrailExecutionStage.POST,
+)
+def answer_question(question: str) -> str:
+    ...
+```
+
+- `guardrail_text` (required) — the rule the judge evaluates against, at most 4000 characters.
+- `model` (required) — the judge model id, e.g. `"gpt-4o-2024-08-06"`. It must be a model your organization's governance policy allows for the LLM-as-judge guardrail — LLM Gateway enforces the permitted list, so a model that isn't authorized for judging is rejected.
+- `threshold` — strictness from `0` (strictest) to `6` (most lenient), defaulting to `2`. Higher values flag only clear violations.
+- `positive_examples` / `negative_examples` — optional example payloads (not descriptions) that comply with / violate the rule, used to calibrate the judge. At most 2 entries per list, each at most 1000 characters.
+
 ## Actions
 
 Actions define what happens when a violation is detected.
@@ -433,6 +467,11 @@ print(result.result, result.reason)
       members:
         - IntellectualPropertyValidator
 
+::: uipath.platform.guardrails.decorators.validators.llm_as_judge
+    options:
+      members:
+        - LLMAsJudgeValidator
+
 ::: uipath.platform.guardrails.decorators.validators.user_prompt_attacks
     options:
       members: