chore: Release 1.7.2 (#3773)

# Changelog

## Improvements:

- **Memory Growth on Performance Evals**: Added `memory_growth_tracking`
as an attribute on `PerformanceEval` to enable additional debug logs for
memory growth.

## Bug Fixes:

- **Gemini 2.5 Metrics:** Fixed Gemini metrics to correctly include
“thinking” tokens.
- **Claude tool calling:** Fixed a bug related to parsing function call
responses when using Claude models

Author: dirkbrnd

cookbook/agent_concepts/tool_concepts/custom_tools/tool_hook_with_state.py

@@ -35,7 +35,7 @@ def grab_customer_profile_hook(
         raise ValueError(f"Customer profile for {cust_id} not found")
     customer_profile = agent.session_state["customer_profiles"][cust_id]
 
-    # Replace the customer_id with the customer_profile
+    # Replace the customer with the customer_profile
     arguments["customer"] = json.dumps(customer_profile)
     # Call the function with the updated arguments
     result = function_call(**arguments)

cookbook/evals/performance/team_response_with_memory_multi_user.py

@@ -1,5 +1,6 @@
 import asyncio
 import random
+import uuid
 
 from agno.agent import Agent
 from agno.eval.performance import PerformanceEval
@@ -8,6 +9,7 @@
 from agno.models.openai import OpenAIChat
 from agno.storage.postgres import PostgresStorage
 from agno.team.team import Team
+from agno.tools.reasoning import ReasoningTools
 
 users = [
     "abel@example.com",
@@ -68,8 +70,8 @@ def get_activities(city: str) -> str:
     return f"The activities in {city} are {', '.join(selected_activities)}."
 
 
-agent_1 = Agent(
-    agent_id="agent_1",
+weather_agent = Agent(
+    agent_id="weather_agent",
     model=OpenAIChat(id="gpt-4o-mini"),
     description="You are a helpful assistant that can answer questions about the weather.",
     instructions="Be concise, reply with one sentence.",
@@ -79,8 +81,8 @@ def get_activities(city: str) -> str:
     add_history_to_messages=True,
 )
 
-agent_2 = Agent(
-    agent_id="agent_2",
+activities_agent = Agent(
+    agent_id="activities_agent",
     model=OpenAIChat(id="gpt-4o-mini"),
     description="You are a helpful assistant that can answer questions about activities in a city.",
     instructions="Be concise, reply with one sentence.",
@@ -91,11 +93,12 @@ def get_activities(city: str) -> str:
 )
 
 team = Team(
-    members=[agent_1, agent_2],
+    members=[weather_agent, activities_agent],
     model=OpenAIChat(id="gpt-4o-mini"),
     instructions="Be concise, reply with one sentence.",
     memory=memory,
     storage=team_storage,
+    tools=[ReasoningTools()],
     markdown=True,
     enable_user_memories=True,
     add_history_to_messages=True,
@@ -108,7 +111,7 @@ async def run_team_for_user(user: str):
         await team.arun(
             message=f"I love {random_city}! What activities and weather can I expect in {random_city}?",
             user_id=user,
-            session_id=f"session_{user}",
+            session_id=f"session_{uuid.uuid4()}",
         )
 
     tasks = []
@@ -132,11 +135,11 @@ async def run_team_for_user(user: str):
     warmup_runs=0,
     measure_runtime=False,
     debug_mode=True,
+    memory_growth_tracking=True,
+    top_n_memory_allocations=10,
 )
 
 if __name__ == "__main__":
     asyncio.run(
-        team_response_with_memory_impact.arun(
-            print_results=True, print_summary=True, with_growth_tracking=True
-        )
+        team_response_with_memory_impact.arun(print_results=True, print_summary=True)
     )

cookbook/evals/performance/team_response_with_memory_simple.py

@@ -85,11 +85,10 @@ async def run_team():
     warmup_runs=0,
     measure_runtime=False,
     debug_mode=True,
+    memory_growth_tracking=True,
 )
 
 if __name__ == "__main__":
     asyncio.run(
-        team_response_with_memory_impact.arun(
-            print_results=True, print_summary=True, with_growth_tracking=True
-        )
+        team_response_with_memory_impact.arun(print_results=True, print_summary=True)
     )

libs/agno/agno/eval/performance.py

@@ -206,8 +206,7 @@ class PerformanceEval:
     # Print detailed results
     print_results: bool = False
     # Print detailed memory growth analysis
-    with_growth_tracking: bool = False
-
+    memory_growth_tracking: bool = False
     # Number of memory allocations to track
     top_n_memory_allocations: int = 5
 
@@ -468,7 +467,7 @@ async def _async_measure_memory_with_growth_tracking(
         return adjusted_usage, current_snapshot
 
     def run(
-        self, *, print_summary: bool = False, print_results: bool = False, with_growth_tracking: bool = False
+        self, *, print_summary: bool = False, print_results: bool = False, memory_growth_tracking: bool = False
     ) -> PerformanceResult:
         """
         Main method to run the performance evaluation.
@@ -537,7 +536,7 @@ def run(
                     live_log.update(status)
 
                     # Measure memory
-                    if self.with_growth_tracking or with_growth_tracking:
+                    if self.memory_growth_tracking or memory_growth_tracking:
                         usage, current_snapshot = self._measure_memory_with_growth_tracking(
                             memory_baseline, previous_snapshot
                         )
@@ -584,7 +583,7 @@ def run(
         return self.result
 
     async def arun(
-        self, *, print_summary: bool = False, print_results: bool = False, with_growth_tracking: bool = False
+        self, *, print_summary: bool = False, print_results: bool = False, memory_growth_tracking: bool = False
     ) -> PerformanceResult:
         """
         Async method to run the performance evaluation of async functions.
@@ -659,7 +658,7 @@ async def arun(
                     live_log.update(status)
 
                     # Measure memory
-                    if self.with_growth_tracking or with_growth_tracking:
+                    if self.memory_growth_tracking or memory_growth_tracking:
                         usage, current_snapshot = await self._async_measure_memory_with_growth_tracking(
                             memory_baseline, previous_snapshot
                         )

libs/agno/agno/memory/v2/db/redis.py

@@ -20,6 +20,7 @@ def __init__(
         port: int = 6379,
         db: int = 0,
         password: Optional[str] = None,
+        ssl: Optional[bool] = False,
         expire: Optional[int] = None,
     ):
         """
@@ -31,6 +32,7 @@ def __init__(
             port (int): Redis port number
             db (int): Redis database number
             password (Optional[str]): Redis password if authentication is required
+            ssl (Optional[bool]): Whether to use SSL for Redis connection
             expire (Optional[int]): TTL (time to live) in seconds for Redis keys. None means no expiration.
         """
         self.prefix = prefix
@@ -41,6 +43,7 @@ def __init__(
             db=db,
             password=password,
             decode_responses=True,  # Automatically decode responses to str
+            ssl=ssl,
         )
         log_debug(f"Created RedisMemoryDb with prefix: '{self.prefix}'")
 

libs/agno/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "agno"
-version = "1.7.1"
+version = "1.7.2"
 description = "Agno: a lightweight library for building Multi-Agent Systems"
 requires-python = ">=3.7,<4"
 readme = "README.md"
@@ -434,4 +434,4 @@ module = [
   "zep_cloud.*",
   "oxylabs.*"
 ]
-ignore_missing_imports = true
+ignore_missing_imports = true

libs/agno/tests/integration/agent/test_user_confirmation_flows.py

@@ -311,7 +311,9 @@ async def get_the_weather(city: str):
             assert response.tools[0].tool_args == {"city": "Tokyo"}
 
             # Mark the tool as confirmed
-            response.tools[0].confirmed = True
+            for tool_response in response.tools:
+                if tool_response.requires_confirmation:
+                    tool_response.confirmed = True
             found_confirmation = True
     assert found_confirmation, "No tools were found to require confirmation"
 

libs/agno/tests/integration/agent/test_user_input_flows.py

@@ -119,7 +119,9 @@ async def get_the_weather(city: str):
     assert response.tools[0].tool_args == {"city": "Tokyo"}
 
     # Provide user input
-    response.tools[0].user_input_schema[0].value = "Tokyo"
+    for tool_response in response.tools:
+        if tool_response.requires_user_input:
+            tool_response.user_input_schema[0].value = "Tokyo"
 
     response = await agent.acontinue_run(response)
     assert response.is_paused is False

libs/agno/tests/integration/teams/test_memory_impact.py

@@ -279,7 +279,7 @@ def simple_function(input_text: str) -> str:
         mode="route",
         model=OpenAIChat(id="gpt-4o-mini"),
         members=[agent],
-        storage=agent_storage,
+        storage=team_storage,
         memory=memory,
         enable_user_memories=True,
     )