Minor fixes

agency_swarm/agents/Devid/tools/util/format_file_deps.py

@@ -39,7 +39,7 @@ def append_dependencies(self):
                     "content": f"Extract the dependencies from the file '{file}'."
                 }
             ],
-            model="gpt-3.5-turbo",
+            model="gpt-4o-mini",
             temperature=0,
             response_format=Dependencies
         )

agency_swarm/util/validators.py

@@ -8,18 +8,9 @@ class Validator(BaseModel):
     Validate if an attribute is correct and if not,
     return a new value with an error message
     """
-    reason: str = Field(
-        default=None,
-        description="Step-by-step reasoning why the attribute could be valid or not with a conclussion at the end.",
-    )
-    is_valid: bool = Field(
-        default=True,
-        description="Whether the attribute is valid based on the requirements.",
-    )
-    fixed_value: str = Field(
-        default=None,
-        description="If the attribute is not valid, suggest a new value for the attribute",
-    )
+    reason: str = Field(..., description="Step-by-step reasoning why the attribute could be valid or not with a conclussion at the end.")
+    is_valid: bool = Field(..., description="Whether the attribute is valid based on the requirements.")
+    fixed_value: str = Field(..., description="If the attribute is not valid, suggest a new value for the attribute. Otherwise, leave it empty.")
 
 def llm_validator(
     statement: str,
@@ -81,6 +72,11 @@ def llm(v: str) -> str:
             temperature=temperature,
         )
 
+        if resp.choices[0].message.refusal:
+            raise ValueError(resp.choices[0].message.refusal)
+
+        resp = resp.choices[0].message.parsed
+
         # If the response is  not valid, return the reason, this could be used in
         # the future to generate a better response, via reasking mechanism.
         assert resp.is_valid, resp.reason