langwatch-scenario 0.7.7__py3-none-any.whl → 0.7.8__py3-none-any.whl

{langwatch_scenario-0.7.7.dist-info → langwatch_scenario-0.7.8.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: langwatch-scenario
-Version: 0.7.7
+Version: 0.7.8
 Summary: The end-to-end agent testing library
 Author-email: LangWatch Team <support@langwatch.ai>
 License: MIT

{langwatch_scenario-0.7.7.dist-info → langwatch_scenario-0.7.8.dist-info}/RECORD

@@ -2,13 +2,13 @@ scenario/__init__.py,sha256=4WO8TjY8Lc0NhYL7b9LvaB1xCBqwUkLuI0uIA6PQP6c,4223
 scenario/_error_messages.py,sha256=QVFSbhzsVNGz2GOBOaoQFW6w6AOyZCWLTt0ySWPfnGw,3882
 scenario/agent_adapter.py,sha256=PoY2KQqYuqzIIb3-nhIU-MPXwHJc1vmwdweMy7ut-hk,4255
 scenario/cache.py,sha256=J6s6Sia_Ce6TrnsInlhfxm6SF8tygo3sH-_cQCRX1WA,6213
-scenario/judge_agent.py,sha256=7NsgeMu6wRMjU_HYTCFqkLma6H2AJuEkw9hJkt11190,16211
+scenario/judge_agent.py,sha256=gWRWzIfHBjAYBRXant6n5fL_E2P3A2IGNvIyp9nUb30,16728
 scenario/pytest_plugin.py,sha256=DGrpgB6e71eq8QXWWxwLjAKNhiyYyzfzZ0L5Ax8iEmo,11317
 scenario/scenario_executor.py,sha256=2ZPy2cywwEMIbUfBP1jHN__Ffjf5WGB144MX2SNr5IM,33101
 scenario/scenario_state.py,sha256=LWGqEQN-Yz0DIiC-TyMRHd-9rEiuBVUHKllMmKv-qGg,7029
 scenario/script.py,sha256=A0N5pP0l4FFn1xdKc78U_wkwWhEWH3EFeU_LRDtNyEI,12241
 scenario/types.py,sha256=qH5KFzJBDG1fEJB_qFRVtL3EZulxq3G1mztYczIzIAY,9613
-scenario/user_simulator_agent.py,sha256=UJ75xhqHwoi8-3JkR1AsHDzpHM2Lx-aDSTJ1gnq_SXc,9101
+scenario/user_simulator_agent.py,sha256=kqnSd4_gytzEwtkc06r58UdE1EycZBzejRPzfORDjdo,9619
 scenario/_events/__init__.py,sha256=4cj6H9zuXzvWhT2P2JNdjWzeF1PUepTjqIDw85Vid9s,1500
 scenario/_events/event_alert_message_logger.py,sha256=K0Pu76Gd36lGEEYh8e8r7NMt7J-OQhbw0cZmiwutCOE,3591
 scenario/_events/event_bus.py,sha256=KFN0OxAQIQXIk_tVrorDoN_YLKVK9dos5SXFALstHgE,9809
@@ -232,10 +232,10 @@ scenario/_utils/message_conversion.py,sha256=AWHn31E7J0mz9sBXWruVVAgtsrJz1R_xEf-
 scenario/_utils/utils.py,sha256=msQgUWaLh3U9jIIHmxkEbOaklga63AF0KJzsaKa_mZc,14008
 scenario/config/__init__.py,sha256=b2X_bqkIrd7jZY9dRrXk2wOqoPe87Nl_SRGuZhlolxA,1123
 scenario/config/langwatch.py,sha256=ijWchFbUsLbQooAZmwyTw4rxfRLQseZ1GoVSiPPbzpw,1677
-scenario/config/model.py,sha256=Ve49S2FyzUifXJ-SAyKPiNtVqs8BfsYbODu_M5y0c8Y,1155
+scenario/config/model.py,sha256=T4HYA79CW1NxXDkFlyftYR6JzZcowbtIx0H-ijxRyfg,1297
 scenario/config/scenario.py,sha256=tVVnsUgG6Z0hYZiTDX-GGZz8l8co1HhyTqJUJNPinBk,5184
-langwatch_scenario-0.7.7.dist-info/METADATA,sha256=L7h0kgOaIij6MYVCac0EqPu8ODkZNKxDeIrHCSJg2l4,20003
-langwatch_scenario-0.7.7.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-langwatch_scenario-0.7.7.dist-info/entry_points.txt,sha256=WlEnJ_gku0i18bIa3DSuGqXRX-QDQLe_s0YmRzK45TI,45
-langwatch_scenario-0.7.7.dist-info/top_level.txt,sha256=45Mn28aedJsetnBMB5xSmrJ-yo701QLH89Zlz4r1clE,9
-langwatch_scenario-0.7.7.dist-info/RECORD,,
+langwatch_scenario-0.7.8.dist-info/METADATA,sha256=q7Rk73qwl5ZzaRTEF9IWxLzgCBniCMO8Ku240jVyBLY,20003
+langwatch_scenario-0.7.8.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+langwatch_scenario-0.7.8.dist-info/entry_points.txt,sha256=WlEnJ_gku0i18bIa3DSuGqXRX-QDQLe_s0YmRzK45TI,45
+langwatch_scenario-0.7.8.dist-info/top_level.txt,sha256=45Mn28aedJsetnBMB5xSmrJ-yo701QLH89Zlz4r1clE,9
+langwatch_scenario-0.7.8.dist-info/RECORD,,

scenario/config/model.py

@@ -17,7 +17,8 @@ class ModelConfig(BaseModel):
     for use with user simulator and judge agents in the Scenario framework.
 
     Attributes:
-        model: The model identifier (e.g., "openai/gpt-4.1-mini", "anthropic/claude-3-sonnet")
+        model: The model identifier (e.g., "openai/gpt-4.1", "anthropic/claude-3-sonnet")
+        api_base: Optional base URL where the model is hosted
         api_key: Optional API key for the model provider
         temperature: Sampling temperature for response generation (0.0 = deterministic, 1.0 = creative)
         max_tokens: Maximum number of tokens to generate in responses
@@ -26,6 +27,7 @@ class ModelConfig(BaseModel):
         ```
         model_config = ModelConfig(
             model="openai/gpt-4.1",
+            api_base="https://api.openai.com/v1",
             api_key="your-api-key",
             temperature=0.1,
             max_tokens=1000
@@ -34,6 +36,7 @@ class ModelConfig(BaseModel):
     """
 
     model: str
+    api_base: Optional[str] = None
     api_key: Optional[str] = None
     temperature: float = 0.0
     max_tokens: Optional[int] = None

scenario/judge_agent.py

@@ -41,6 +41,7 @@ class JudgeAgent(AgentAdapter):
     Attributes:
         role: Always AgentRole.JUDGE for judge agents
         model: LLM model identifier to use for evaluation
+        api_base: Optional base URL where the model is hosted
         api_key: Optional API key for the model provider
         temperature: Sampling temperature for evaluation consistency
         max_tokens: Maximum tokens for judge reasoning
@@ -97,6 +98,7 @@ class JudgeAgent(AgentAdapter):
     role = AgentRole.JUDGE
 
     model: str
+    api_base: Optional[str]
     api_key: Optional[str]
     temperature: float
     max_tokens: Optional[int]
@@ -108,6 +110,7 @@ class JudgeAgent(AgentAdapter):
         *,
         criteria: Optional[List[str]] = None,
         model: Optional[str] = None,
+        api_base: Optional[str] = None,
         api_key: Optional[str] = None,
         temperature: float = 0.0,
         max_tokens: Optional[int] = None,
@@ -122,6 +125,8 @@ class JudgeAgent(AgentAdapter):
                      and negative constraints ("Agent should not provide personal information").
             model: LLM model identifier (e.g., "openai/gpt-4.1").
                    If not provided, uses the default model from global configuration.
+            api_base: Optional base URL where the model is hosted. If not provided,
+                      uses the base URL from global configuration.
             api_key: API key for the model provider. If not provided,
                      uses the key from global configuration or environment.
             temperature: Sampling temperature for evaluation (0.0-1.0).
@@ -156,6 +161,7 @@ class JudgeAgent(AgentAdapter):
         """
         # Override the default system prompt for the judge agent
         self.criteria = criteria or []
+        self.api_base = api_base
         self.api_key = api_key
         self.temperature = temperature
         self.max_tokens = max_tokens
@@ -172,6 +178,9 @@ class JudgeAgent(AgentAdapter):
             ScenarioConfig.default_config.default_model, ModelConfig
         ):
             self.model = model or ScenarioConfig.default_config.default_model.model
+            self.api_base = (
+                api_base or ScenarioConfig.default_config.default_model.api_base
+            )
             self.api_key = (
                 api_key or ScenarioConfig.default_config.default_model.api_key
             )
@@ -351,6 +360,8 @@ if you don't have enough information to make a verdict, say inconclusive with ma
                 model=self.model,
                 messages=messages,
                 temperature=self.temperature,
+                api_key=self.api_key,
+                api_base=self.api_base,
                 max_tokens=self.max_tokens,
                 tools=tools,
                 tool_choice=(

scenario/user_simulator_agent.py

@@ -37,6 +37,7 @@ class UserSimulatorAgent(AgentAdapter):
     Attributes:
         role: Always AgentRole.USER for user simulator agents
         model: LLM model identifier to use for generating user messages
+        api_base: Optional base URL where the model is hosted
         api_key: Optional API key for the model provider
         temperature: Sampling temperature for response generation
         max_tokens: Maximum tokens to generate in user messages
@@ -76,9 +77,11 @@ class UserSimulatorAgent(AgentAdapter):
         - Messages are generated in a casual, human-like style (lowercase, brief, etc.)
         - The simulator will not act as an assistant - it only generates user inputs
     """
+
     role = AgentRole.USER
 
     model: str
+    api_base: Optional[str]
     api_key: Optional[str]
     temperature: float
     max_tokens: Optional[int]
@@ -88,6 +91,7 @@ class UserSimulatorAgent(AgentAdapter):
         self,
         *,
         model: Optional[str] = None,
+        api_base: Optional[str] = None,
         api_key: Optional[str] = None,
         temperature: float = 0.0,
         max_tokens: Optional[int] = None,
@@ -99,6 +103,8 @@ class UserSimulatorAgent(AgentAdapter):
         Args:
             model: LLM model identifier (e.g., "openai/gpt-4.1").
                    If not provided, uses the default model from global configuration.
+            api_base: Optional base URL where the model is hosted. If not provided,
+                      uses the base URL from global configuration.
             api_key: API key for the model provider. If not provided,
                      uses the key from global configuration or environment.
             temperature: Sampling temperature for message generation (0.0-1.0).
@@ -128,6 +134,7 @@ class UserSimulatorAgent(AgentAdapter):
             ```
         """
         # Override the default system prompt for the user simulator agent
+        self.api_base = api_base
         self.api_key = api_key
         self.temperature = temperature
         self.max_tokens = max_tokens
@@ -144,6 +151,9 @@ class UserSimulatorAgent(AgentAdapter):
             ScenarioConfig.default_config.default_model, ModelConfig
         ):
             self.model = model or ScenarioConfig.default_config.default_model.model
+            self.api_base = (
+                api_base or ScenarioConfig.default_config.default_model.api_base
+            )
             self.api_key = (
                 api_key or ScenarioConfig.default_config.default_model.api_key
             )
@@ -222,6 +232,8 @@ Your goal (assistant) is to interact with the Agent Under Test (user) as if you
                 model=self.model,
                 messages=messages,
                 temperature=self.temperature,
+                api_key=self.api_key,
+                api_base=self.api_base,
                 max_tokens=self.max_tokens,
                 tools=[],
             ),