feat: Add created_at stamp for chat messages to allow agents and conversational app to reason over current time

Author: YoanSallami

coverage-badge.svg

@@ -309,6 +309,7 @@ class Query(synalinks.DataModel):
 
 class SQLResult(synalinks.DataModel):
     """The result of the SQL agent's analysis."""
+
     answer: str = synalinks.Field(
         description="A clear, natural language answer to the user's question"
     )
@@ -454,10 +455,7 @@ async def populate_knowledge_base(kb):
     # Sample customers
     customers = [
         Customer(
-            id="C001",
-            name="Alice Johnson",
-            email="alice@example.com",
-            country="USA"
+            id="C001", name="Alice Johnson", email="alice@example.com", country="USA"
         ),
         Customer(
             id="C002",
@@ -537,7 +535,7 @@ async def populate_knowledge_base(kb):
             stock=120,
         ),
         Product(
-            id="P008", 
+            id="P008",
             name="Notebook Set",
             category="Office",
             price=12.99,
@@ -738,7 +736,9 @@ async def main():
                         tool_calls_count += 1
                         args = tool_call.get("arguments", {})
                         args_str = ", ".join(f"{k}={repr(v)}" for k, v in args.items())
-                        print(f"Tool Call {tool_calls_count}: {tool_call['name']}({args_str})")
+                        print(
+                            f"Tool Call {tool_calls_count}: {tool_call['name']}({args_str})"
+                        )
                 elif msg.get("role") == "tool":
                     content = msg.get("content", "")
                     # Truncate long results for readability

examples/16_sql_agent.py

@@ -4,7 +4,7 @@
 ## Understanding Rewards
 
 `Reward`s are an essential part of reinforcement learning frameworks.
-They are scalar values (between 0.0 and 1.0 for synalinks) 
+They are scalar values (between 0.0 and 1.0 for synalinks)
 that guide the process into making more efficient decisions or
 predictions. During training, the goal is to maximize the reward function.
 The reward gives the system an indication of how well it performed for that task.

examples/7_reward_metrics_and_optimizers.py

@@ -202,7 +202,6 @@ async def main():
 
 import synalinks
 
-
 # =============================================================================
 # Data Models
 # =============================================================================
@@ -331,7 +330,7 @@ async def main():
     print("=" * 60)
 
     result = await program(Query(query="What is the capital of France?"))
-    print(f"\nQuery: 'What is the capital of France?'")
+    print("\nQuery: 'What is the capital of France?'")
     print(f"Result: {result.get_json()}")
 
     # -------------------------------------------------------------------------
@@ -342,7 +341,7 @@ async def main():
     print("=" * 60)
 
     result = await program(Query(query="Explain what Python is used for"))
-    print(f"\nQuery: 'Explain what Python is used for'")
+    print("\nQuery: 'Explain what Python is used for'")
     print(f"Result: {result.get_json()}")
 
     # -------------------------------------------------------------------------

guides/10_output_guard.py

@@ -246,12 +246,8 @@ class Question(synalinks.DataModel):
 class Answer(synalinks.DataModel):
     """Output: An answer with reasoning."""
 
-    thinking: str = synalinks.Field(
-        description="Your step-by-step reasoning process"
-    )
-    answer: str = synalinks.Field(
-        description="The final answer based on your reasoning"
-    )
+    thinking: str = synalinks.Field(description="Your step-by-step reasoning process")
+    answer: str = synalinks.Field(description="The final answer based on your reasoning")
 
 
 async def main():

guides/1_getting_started.py

@@ -108,7 +108,8 @@ class MovieReview(synalinks.DataModel):
 
 ### Using Enums for Constrained Outputs
 
-When you need the LLM to choose from specific options, use Python Enums (similar to the Literal above):
+When you need the LLM to choose from specific options,
+use Python Enums (similar to the Literal above):
 
 ```python
 from enum import Enum

guides/2_data_models.py

@@ -330,7 +330,6 @@ async def build(self, inputs: synalinks.SymbolicDataModel) -> None:
 
 import synalinks
 
-
 # =============================================================================
 # Data Models
 # =============================================================================

guides/3_programs.py

@@ -345,7 +345,6 @@ async def main():
 
 import synalinks
 
-
 # =============================================================================
 # Data Models
 # =============================================================================

guides/6_knowledge_base.py

@@ -249,7 +249,10 @@ async def main():
         (MathProblem(problem="5 * 4"), MathAnswer(thinking="5 * 4 = 20", answer="20")),
         (MathProblem(problem="10 - 3"), MathAnswer(thinking="10 - 3 = 7", answer="7")),
         (MathProblem(problem="8 / 2"), MathAnswer(thinking="8 / 2 = 4", answer="4")),
-        (MathProblem(problem="3 + 3 + 3"), MathAnswer(thinking="3 + 3 + 3 = 9", answer="9")),
+        (
+            MathProblem(problem="3 + 3 + 3"),
+            MathAnswer(thinking="3 + 3 + 3 = 9", answer="9"),
+        ),
         (MathProblem(problem="7 * 2"), MathAnswer(thinking="7 * 2 = 14", answer="14")),
     ]
 
@@ -398,7 +401,6 @@ async def main():
 
 import synalinks
 
-
 # =============================================================================
 # Data Models
 # =============================================================================
@@ -478,7 +480,7 @@ async def main():
         ],
         dtype="object",
     )
-    y_test = np.array(
+    _y_test = np.array(  # noqa: F841
         [
             MathAnswer(thinking="4 + 5 = 9", answer="9"),
             MathAnswer(thinking="6 * 3 = 18", answer="18"),

guides/7_training.py

@@ -308,7 +308,6 @@ async def main():
 
 import synalinks
 
-
 # =============================================================================
 # Data Models
 # =============================================================================