chore: Release 1.7.1 (#3740)

dirkbrnd · kausmeows · web-flow · commit e9d640f53cf3 · 2025-07-04T10:11:10.000+02:00
# Changelog ## New Features: - **Debug Level:** Added `debug_level` to both `Agent` and `Team`. This is an `int` value of either `1` (default) or `2` (more verbose logging). Currently this only enables more verbose model logs, but will be used more widely in future. ## Improvements: - **Parser Model on Teams**: Added `parser_model` support for `Team`. See docs [here](https://docs.agno.com/teams/structured-output#using-a-parser-model). - **Support for Gemini Thinking**: Added `thinking_budget` and `include_thoughts` parameters for `Gemini` model class. - **Serper Tools**: Made updates to the toolkit to include new tools `search_news` , `search_scholar` and `scrape_webpage`. - **Valyu Tools:** New Valyu toolkit for Deep Search capabilities of academic sources. - **Oxylabs:** Added `OxylabsTools` for adding more web-scraping capabilities to agents. ## Bug Fixes: - **DuckDB CSV parsing error:** For CSV files use the custom `read_csv` method for improved CSV parsing - **Full Team Metrics**: Fixed an issue with calculation of the `full_team_session_metrics` on Teams. --------- Co-authored-by: Kaustubh <shuklakaustubh84@gmail.com>
diff --git a/.github/workflows/test_on_release.yml b/.github/workflows/test_on_release.yml
@@ -629,6 +629,7 @@ jobs:
     env:
       OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
       ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
+      GOOGLE_API_KEY: ${{ secrets.GOOGLE_API_KEY }}
     steps:
       - uses: actions/checkout@v3
       - name: Set up Python ${{ matrix.python-version }}
diff --git a/README.md b/README.md
@@ -226,13 +226,19 @@ Agno agents are designed for performance and while we do share some benchmarks a
 
 Given that each framework is different and we won't be able to tune their performance like we do with Agno, for future benchmarks we'll only be comparing against ourselves.
 
-## Cursor Setup
+## Complete Documentation Index
+
+For LLMs and AI assistants to understand and navigate Agno's complete documentation, we provide an [LLMs.txt](https://docs.agno.com/llms.txt) or [LLMs-Full.txt](https://docs.agno.com/llms-full.txt) file.
+
+This file is specifically formatted for AI systems to efficiently parse and reference our documentation.
+
+### Cursor Setup
 
 When building Agno agents, using Agno documentation as a source in Cursor is a great way to speed up your development.
 
-1. In Cursor, go to the settings or preferences section.
-2. Find the section to manage documentation sources.
-3. Add `https://docs.agno.com` to the list of documentation URLs.
+1. In Cursor, go to the "Cursor Settings" menu.
+2. Find the "Indexing & Docs" section.
+3. Add `https://docs.agno.com/llms-full.txt` to the list of documentation URLs.
 4. Save the changes.
 
 Now, Cursor will have access to the Agno documentation.
diff --git a/cookbook/examples/streamlit_apps/agentic_rag/utils.py b/cookbook/examples/streamlit_apps/agentic_rag/utils.py
@@ -94,7 +94,7 @@ def display_tool_calls(tool_calls_container, tools: List[ToolExecution]):
                 if _metrics:
                     st.markdown("**Metrics:**")
                     st.json(
-                        _metrics if isinstance(_metrics, dict) else _metrics._to_dict()
+                        _metrics if isinstance(_metrics, dict) else _metrics.to_dict()
                     )
 
 
diff --git a/cookbook/examples/streamlit_apps/parallel_world_builder/utils.py b/cookbook/examples/streamlit_apps/parallel_world_builder/utils.py
@@ -98,5 +98,5 @@ def display_tool_calls(tool_calls_container, tools: List[ToolExecution]):
                 if _metrics:
                     st.markdown("**Metrics:**")
                     st.json(
-                        _metrics if isinstance(_metrics, dict) else _metrics._to_dict()
+                        _metrics if isinstance(_metrics, dict) else _metrics.to_dict()
                     )
diff --git a/libs/agno/agno/agent/agent.py b/libs/agno/agno/agent/agent.py
@@ -1243,9 +1243,10 @@ async def _arun_stream(
             yield event
 
         # If a parser model is provided, structure the response separately
-        await self._aparse_response_with_parser_model_stream(
+        async for event in self._aparse_response_with_parser_model_stream(
             run_response=run_response, stream_intermediate_steps=stream_intermediate_steps
-        )
+        ):
+            yield event
 
         # 3. Add the run to memory
         self._add_run_to_memory(
@@ -7381,7 +7382,9 @@ async def aprint_response(
                     # Check if we have any response content to display
                     response_content = (
                         response_content_stream
-                        if response_content_stream and isinstance(response_content_stream, str) and len(response_content_stream) > 0
+                        if response_content_stream
+                        and isinstance(response_content_stream, str)
+                        and len(response_content_stream) > 0
                         else response_content_batch
                     )
                     if response_content:
diff --git a/libs/agno/agno/embedder/fastembed.py b/libs/agno/agno/embedder/fastembed.py
@@ -1,13 +1,14 @@
 from dataclasses import dataclass
 from typing import Dict, List, Optional, Tuple
+
 from agno.embedder.base import Embedder
 from agno.utils.log import logger
 
 try:
     import numpy as np
 
 except ImportError:
-    raise ImportError("numpy not installed, use pip install numpy")
+    raise ImportError("numpy not installed, use `pip install numpy`")
 
 
 try:
diff --git a/libs/agno/agno/embedder/sentence_transformer.py b/libs/agno/agno/embedder/sentence_transformer.py
@@ -10,6 +10,12 @@
 except ImportError:
     raise ImportError("`sentence-transformers` not installed, please run `pip install sentence-transformers`")
 
+try:
+    import numpy as np
+
+except ImportError:
+    raise ImportError("numpy not installed, use `pip install numpy`")
+
 
 @dataclass
 class SentenceTransformerEmbedder(Embedder):
@@ -26,6 +32,9 @@ def get_embedding(self, text: Union[str, List[str]]) -> List[float]:
             model = self.sentence_transformer_client
         embedding = model.encode(text, prompt=self.prompt, normalize_embeddings=self.normalize_embeddings)
         try:
+            if isinstance(embedding, np.ndarray):
+                return embedding.tolist()
+
             return embedding  # type: ignore
         except Exception as e:
             logger.warning(e)
diff --git a/libs/agno/agno/models/message.py b/libs/agno/agno/models/message.py
@@ -73,7 +73,7 @@ class MessageMetrics:
 
     timer: Optional[Timer] = None
 
-    def _to_dict(self) -> Dict[str, Any]:
+    def to_dict(self) -> Dict[str, Any]:
         metrics_dict = asdict(self)
         metrics_dict.pop("timer")
         metrics_dict = {
@@ -267,7 +267,7 @@ def to_dict(self) -> Dict[str, Any]:
         if self.references:
             message_dict["references"] = self.references.model_dump()
         if self.metrics:
-            message_dict["metrics"] = self.metrics._to_dict()
+            message_dict["metrics"] = self.metrics.to_dict()
             if not message_dict["metrics"]:
                 message_dict.pop("metrics")
 
diff --git a/libs/agno/agno/models/response.py b/libs/agno/agno/models/response.py
@@ -50,7 +50,7 @@ def is_paused(self) -> bool:
     def to_dict(self) -> Dict[str, Any]:
         _dict = asdict(self)
         if self.metrics is not None:
-            _dict["metrics"] = self.metrics._to_dict()
+            _dict["metrics"] = self.metrics.to_dict()
 
         if self.user_input_schema is not None:
             _dict["user_input_schema"] = [field.to_dict() for field in self.user_input_schema]
diff --git a/libs/agno/agno/team/team.py b/libs/agno/agno/team/team.py
@@ -1445,9 +1445,10 @@ async def _arun_stream(
             yield event
 
         # If a parser model is provided, structure the response separately
-        await self._aparse_response_with_parser_model_stream(
+        async for event in self._aparse_response_with_parser_model_stream(
             run_response=run_response, stream_intermediate_steps=stream_intermediate_steps
-        )
+        ):
+            yield event
 
         # 3. Add the run to memory
         self._add_run_to_memory(
diff --git a/libs/agno/agno/tools/mcp.py b/libs/agno/agno/tools/mcp.py
@@ -83,7 +83,7 @@ def __init__(
         super().__init__(name="MCPTools", **kwargs)
 
         # Set these after `__init__` to bypass the `_check_tools_filters`
-        # beacuse tools are not available until `initialize()` is called.
+        # because tools are not available until `initialize()` is called.
         self.include_tools = include_tools
         self.exclude_tools = exclude_tools
 
diff --git a/libs/agno/agno/utils/gemini.py b/libs/agno/agno/utils/gemini.py
@@ -156,7 +156,7 @@ def convert_schema(schema_dict: Dict[str, Any], root_schema: Optional[Dict[str,
                     )
                 }
                 if value_type == "ARRAY":
-                    placeholder_properties["example_key"].items = {}
+                    placeholder_properties["example_key"].items = {}  # type: ignore
 
                 return Schema(
                     type=Type.OBJECT,
diff --git a/libs/agno/agno/utils/timer.py b/libs/agno/agno/utils/timer.py
@@ -32,3 +32,10 @@ def __exit__(self, *args):
         self.end_time = perf_counter()
         if self.start_time is not None:
             self.elapsed_time = self.end_time - self.start_time
+
+    def to_dict(self):
+        return {
+            "start_time": str(self.start_time) if self.start_time is not None else None,
+            "end_time": str(self.end_time) if self.end_time is not None else None,
+            "elapsed": self.elapsed,
+        }
diff --git a/libs/agno/agno/vectordb/qdrant/qdrant.py b/libs/agno/agno/vectordb/qdrant/qdrant.py
@@ -451,7 +451,7 @@ def search(self, query: str, limit: int = 5, filters: Optional[Dict[str, Any]] =
             limit (int): Number of search results to return
             filters (Optional[Dict[str, Any]]): Filters to apply while searching
         """
-        filters = self._format_filters(filters or {})
+        filters = self._format_filters(filters or {})  # type: ignore
         if self.search_type == SearchType.vector:
             results = self._run_vector_search_sync(query, limit, filters)
         elif self.search_type == SearchType.keyword:
@@ -466,7 +466,7 @@ def search(self, query: str, limit: int = 5, filters: Optional[Dict[str, Any]] =
     async def async_search(
         self, query: str, limit: int = 5, filters: Optional[Dict[str, Any]] = None
     ) -> List[Document]:
-        filters = self._format_filters(filters or {})
+        filters = self._format_filters(filters or {})  # type: ignore
         if self.search_type == SearchType.vector:
             results = await self._run_vector_search_async(query, limit, filters)
         elif self.search_type == SearchType.keyword:
diff --git a/libs/agno/pyproject.toml b/libs/agno/pyproject.toml
@@ -1,6 +1,6 @@
 [project]
 name = "agno"
-version = "1.7.0"
+version = "1.7.1"
 description = "Agno: a lightweight library for building Multi-Agent Systems"
 requires-python = ">=3.7,<4"
 readme = "README.md"
diff --git a/libs/agno/tests/integration/agent/test_user_confirmation_flows.py b/libs/agno/tests/integration/agent/test_user_confirmation_flows.py
@@ -1,3 +1,4 @@
+
 import pytest
 
 from agno.agent import Agent, RunResponse  # noqa
@@ -286,7 +287,6 @@ async def get_the_weather(city: str):
     response.tools[0].confirmed = True
 
     response = await agent.acontinue_run(response)
-    assert response.is_paused is False
     assert response.tools[0].result == "It is currently 70 degrees and cloudy in Tokyo"
 
 
diff --git a/libs/agno/tests/integration/teams/test_team_with_storage_and_memory.py b/libs/agno/tests/integration/teams/test_team_with_storage_and_memory.py
@@ -8,6 +8,7 @@
 from agno.memory.v2.db.sqlite import SqliteMemoryDb
 from agno.memory.v2.memory import Memory
 from agno.models.anthropic.claude import Claude
+from agno.models.google.gemini import Gemini
 from agno.models.openai.chat import OpenAIChat
 from agno.storage.sqlite import SqliteStorage
 from agno.team.team import Team
@@ -77,6 +78,37 @@ def route_team(team_storage, memory):
     )
 
 
+@pytest.fixture
+def route_team_with_members(team_storage, agent_storage, memory):
+    """Create a route team with storage and memory for testing."""
+
+    def get_weather(city: str) -> str:
+        return f"The weather in {city} is sunny."
+
+    def get_open_restaurants(city: str) -> str:
+        return f"The open restaurants in {city} are: {', '.join(['Restaurant 1', 'Restaurant 2', 'Restaurant 3'])}"
+
+    travel_agent = Agent(
+        name="Travel Agent",
+        model=Gemini(id="gemini-2.0-flash-001"),
+        storage=agent_storage,
+        memory=memory,
+        add_history_to_messages=True,
+        role="Search the web for travel information. Don't call multiple tools at once. First get weather, then restaurants.",
+        tools=[get_weather, get_open_restaurants],
+    )
+    return Team(
+        name="Route Team",
+        mode="route",
+        model=Gemini(id="gemini-2.0-flash-001"),
+        members=[travel_agent],
+        storage=team_storage,
+        memory=memory,
+        instructions="Route a single question to the travel agent. Don't make multiple requests.",
+        enable_user_memories=True,
+    )
+
+
 @pytest.mark.asyncio
 async def test_run_history_persistence(route_team, team_storage, memory):
     """Test that all runs within a session are persisted in storage."""
@@ -142,6 +174,65 @@ async def test_run_session_summary(route_team, team_storage, memory):
     assert len(team_session.memory["summaries"][user_id][session_id]) > 0
 
 
+@pytest.mark.asyncio
+async def test_member_run_history_persistence(route_team_with_members, agent_storage, memory):
+    """Test that all runs within a member's session are persisted in storage."""
+    user_id = "john@example.com"
+    session_id = "session_123"
+
+    # Clear memory for this specific test case
+    memory.clear()
+
+    # First request
+    await route_team_with_members.arun(
+        "I'm traveling to Tokyo, what is the weather and open restaurants?", user_id=user_id, session_id=session_id
+    )
+
+    agent_session = agent_storage.read(session_id=session_id)
+
+    stored_memory_data = agent_session.memory
+    assert stored_memory_data is not None, "Memory data not found in stored session."
+
+    agent_runs = stored_memory_data["runs"]
+
+    assert len(agent_runs[-1]["messages"]) == 7, (
+        "Only system message, user message, two tool calls (and results), and response"
+    )
+
+    first_user_message_content = agent_runs[0]["messages"][1]["content"]
+    assert "I'm traveling to Tokyo, what is the weather and open restaurants?" in first_user_message_content
+
+    # Second request
+    await route_team_with_members.arun(
+        "I'm traveling to Munich, what is the weather and open restaurants?", user_id=user_id, session_id=session_id
+    )
+
+    agent_session = agent_storage.read(session_id=session_id)
+
+    stored_memory_data = agent_session.memory
+    assert stored_memory_data is not None, "Memory data not found in stored session."
+
+    agent_runs = stored_memory_data["runs"]
+
+    assert len(agent_runs[-1]["messages"]) == 13, "Full history of messages"
+
+    # Third request (to the member directly)
+    await route_team_with_members.members[0].arun(
+        "Write me a report about all the places I have requested information about",
+        user_id=user_id,
+        session_id=session_id,
+    )
+
+    agent_session = agent_storage.read(session_id=session_id)
+
+    stored_memory_data = agent_session.memory
+    assert stored_memory_data is not None, "Memory data not found in stored session."
+
+    agent_runs = stored_memory_data["runs"]
+
+    assert len(agent_runs[-1]["messages"]) == 15, "Full history of messages"
+
+
 @pytest.mark.asyncio
 async def test_multi_user_multi_session_route_team(route_team, team_storage, memory):
     """Test multi-user multi-session route team with storage and memory."""
diff --git a/libs/agno/tests/unit/response.py b/libs/agno/tests/unit/response.py
@@ -0,0 +1,16 @@
+import json
+
+from agno.models.message import Message, MessageMetrics
+from agno.run.response import RunResponse
+
+
+def test_timer_serialization():
+    message_1 = Message(role="user", content="Hello, world!")
+    message_2 = Message(role="assistant", metrics=MessageMetrics())
+
+    message_2.metrics.start_timer()
+    message_2.metrics.stop_timer()
+
+    run_response = RunResponse(messages=[message_1, message_2])
+
+    assert json.dumps(run_response.to_dict()) is not None
diff --git a/libs/agno/tests/unit/tools/test_oxylabs.py b/libs/agno/tests/unit/tools/test_oxylabs.py

Original file line number	Diff line number	Diff line change
`@@ -94,7 +94,7 @@ def display_tool_calls(tool_calls_container, tools: List[ToolExecution]):`
`94`	`94`	`if _metrics:`
`95`	`95`	`st.markdown("Metrics:")`
`96`	`96`	`st.json(`
`97`		`- _metrics if isinstance(_metrics, dict) else _metrics._to_dict()`
	`97`	`+ _metrics if isinstance(_metrics, dict) else _metrics.to_dict()`
`98`	`98`	`)`
`99`	`99`
`100`	`100`
Original file line number	Diff line number	Diff line change
`@@ -98,5 +98,5 @@ def display_tool_calls(tool_calls_container, tools: List[ToolExecution]):`
`98`	`98`	`if _metrics:`
`99`	`99`	`st.markdown("Metrics:")`
`100`	`100`	`st.json(`
`101`		`- _metrics if isinstance(_metrics, dict) else _metrics._to_dict()`
	`101`	`+ _metrics if isinstance(_metrics, dict) else _metrics.to_dict()`
`102`	`102`	`)`