Managing Tool Calls

This example demonstrates how to use max_tool_calls_from_history to limit tool calls in team context across multiple research queries.

Code

filter_tool_calls_from_history.py

from textwrap import dedent

from agno.agent import Agent
from agno.db.sqlite import SqliteDb
from agno.models.openai import OpenAIResponses
from agno.team import Team
from agno.tools.hackernews import HackerNewsTools
from agno.tools.yfinance import YFinanceTools

# Create specialized research agents
tech_researcher = Agent(
    name="Alex",
    role="Technology Researcher",
    instructions=dedent("""
        You specialize in technology and AI research.
        - Focus on latest developments, trends, and breakthroughs
        - Provide concise, data-driven insights
    """).strip(),
)

business_analyst = Agent(
    name="Sarah",
    role="Business Analyst",
    instructions=dedent("""
        You specialize in business and market analysis.
        - Focus on companies, markets, and economic trends
        - Provide actionable business insights
        - Include relevant data and statistics
    """).strip(),
)

# Create research team with tools and context management
research_team = Team(
    name="Research Team",
    model=OpenAIResponses(id="gpt-5.2"),
    members=[tech_researcher, business_analyst],
    tools=[HackerNewsTools(), YFinanceTools()],
    description="Research team that investigates topics and provides analysis.",
    instructions=dedent("""
        You are a research coordinator that investigates topics comprehensively.

        Your Process:
        1. Use HackerNews to find tech discussions and YFinance for market data
        2. Delegate detailed analysis to the appropriate specialist
        3. Synthesize research findings with specialist insights

        Guidelines:
        - Use HackerNews for tech news and YFinance for financial data
        - Choose the right specialist based on the topic (tech vs business)
        - Combine your research with specialist analysis
        - Provide comprehensive responses
    """).strip(),
    db=SqliteDb(db_file="tmp/research_team.db"),
    session_id="research_session",
    add_history_to_context=True,
    num_history_runs=6,  # Load last 6 research queries
    max_tool_calls_from_history=3,  # Keep only last 3 research results
    markdown=True,
)

research_team.print_response("What are the latest developments in AI agents?", stream=True)
research_team.print_response("How is NVDA performing this quarter?", stream=True)
research_team.print_response("What are the trends in LLM applications?", stream=True)
research_team.print_response("What companies are leading in AI infrastructure?", stream=True)

Usage

1. Create a Python file

   Create filter_tool_calls_from_history.py with the code above.

2. Set up your virtual environment

   Mac

   uv venv --python 3.12
   source .venv/bin/activate
   

   uv venv --python 3.12
   .venv\Scripts\activate
   
3. Install dependencies

   uv pip install -U agno openai yfinance sqlalchemy
   
4. Export your OpenAI API key

   Mac/Linux

   export OPENAI_API_KEY="your_openai_api_key_here"
   

   $Env:OPENAI_API_KEY="your_openai_api_key_here"
   
5. Run Team

   python filter_tool_calls_from_history.py