Added a configuration flag to set the maximum number of tokens in validation prompts

tests/test_llm_validators.py

@@ -1092,7 +1092,10 @@ def test_llm_validators_use_shared_token_batching_helper(monkeypatch, tmp_path):
     assert [decision.approved for decision in result.decisions] == [True, True, True]
     assert len(client.calls) == 2
     assert len(chunk_calls) == 1
-    assert chunk_calls[0]["max_tokens"] == AuditaConfig.from_sources(env={"OPENROUTER_API_KEY": "test-key"}).max_section_tokens
+    assert (
+        chunk_calls[0]["max_tokens"]
+        == AuditaConfig.from_sources(env={"OPENROUTER_API_KEY": "test-key"}).validation_max_prompt_tokens
+    )
     assert all("corrected_segment_text" in payload for payload in chunk_calls[0]["payloads"])
 
 
@@ -1272,6 +1275,74 @@ def test_validator_uses_validation_llm_config(tmp_path):
     assert validation_config.llm_concurrency == 3
 
 
+def test_validator_batches_use_validation_max_prompt_tokens(tmp_path, monkeypatch):
+    transcript = parse_transcript_json(
+        """
+        [
+          {"id": 1, "speaker": "A", "start": 0.0, "end": 1.0, "text": "There were gestures at the temple."}
+        ]
+        """
+    )
+    proposals = [
+        CorrectionProposal(
+            proposal_index=0,
+            module_instance="homophones",
+            module_key="homophones",
+            id=1,
+            original_text="gestures",
+            corrected_text="Jesters",
+            confidence=0.95,
+        )
+    ]
+    client = FakeStructuredLLMClient(
+        [
+            {
+                "validations": [
+                    {
+                        "correction_index": 0,
+                        "approved": True,
+                        "confidence": 0.97,
+                        "reason": "ok",
+                    }
+                ]
+            }
+        ]
+    )
+    chunk_calls = []
+
+    def fake_chunk_payload_items(items, max_tokens, payload_fn, empty_error_message):
+        chunk_calls.append(
+            {
+                "count": len(items),
+                "max_tokens": max_tokens,
+            }
+        )
+        return [TokenBatch(batch_index=0, items=list(items), token_count=1)]
+
+    monkeypatch.setattr(llm_module, "chunk_payload_items", fake_chunk_payload_items)
+    config = AuditaConfig.from_sources(
+        env={"OPENROUTER_API_KEY": "test-key"},
+        overrides=ConfigOverrides(validation_max_prompt_tokens=1024),
+    )
+    context = ValidationContext(
+        proposals=proposals,
+        transcript=transcript,
+        glossary=_glossary(),
+        config=config.validation_llm_config(),
+        run_spec=ModuleRunSpec(
+            instance_name="homophones",
+            module_key="homophones",
+            module=_Module("require_unique"),
+        ),
+        run_dir=tmp_path,
+        llm_client=client,
+    )
+
+    SpokenFormPlausibilityValidator("spoken_form_plausibility_review").validate(context)
+
+    assert chunk_calls[0]["max_tokens"] == 1024
+
+
 def test_validator_uses_validation_llm_concurrency_override(tmp_path, monkeypatch):
     transcript = parse_transcript_json(
         """