/a2a/teams/{id}, not the Agent namespace.
team.py
"""
Serve a Team over A2A
=====================
Expose a research Team through the A2A team namespace. Its card and message
routes live under `/a2a/teams/{id}`, not the Agent namespace.
Prerequisites: OPENAI_API_KEY and the `agno[a2a]` extra
Run: .venvs/demo/bin/python cookbook/05_agent_os/15_a2a/team.py
Try: With the server running, rerun this file with --demo to call POST http://127.0.0.1:7779/a2a/teams/research-team/v1/message:send
"""
import asyncio
import sys
from agno.agent import Agent
from agno.client.a2a import A2AClient
from agno.db.sqlite import SqliteDb
from agno.models.openai import OpenAIResponses
from agno.os import AgentOS
from agno.team import Team
# ---------------------------------------------------------------------------
# Create Research Team
# ---------------------------------------------------------------------------
TEAM_URL = "http://127.0.0.1:7779/a2a/teams/research-team"
db = SqliteDb(
id="a2a-team-db",
db_file="tmp/a2a_team.db",
)
researcher = Agent(
id="a2a-researcher",
name="Researcher",
role="Identify the important facts and tradeoffs.",
model=OpenAIResponses(id="gpt-5.5"),
db=db,
instructions="Analyze the request and return concise factual notes.",
)
writer = Agent(
id="a2a-writer",
name="Writer",
role="Turn research notes into a clear response.",
model=OpenAIResponses(id="gpt-5.5"),
db=db,
instructions="Write a compact answer from the available research.",
)
research_team = Team(
id="research-team",
name="Research Team",
description="A two-member team that researches and writes concise answers.",
model=OpenAIResponses(id="gpt-5.5"),
members=[researcher, writer],
db=db,
instructions=[
"Delegate fact finding to the Researcher.",
"Delegate final wording to the Writer.",
"Return one concise combined answer.",
],
show_members_responses=True,
markdown=True,
)
# ---------------------------------------------------------------------------
# Create AgentOS
# ---------------------------------------------------------------------------
agent_os = AgentOS(
id="a2a-team-os",
description="AgentOS exposing a Team through A2A.",
teams=[research_team],
a2a_interface=True,
)
app = agent_os.get_app()
async def run_demo() -> None:
"""Discover and call the Team through its entity-scoped A2A URL."""
client = A2AClient(TEAM_URL, timeout=90)
card = await client.aget_agent_card()
if card is None:
raise RuntimeError(f"No Team card found at {TEAM_URL}")
result = await client.send_message(
"Give two practical reasons to keep API examples small."
)
if not result.is_completed:
raise RuntimeError(f"Team task ended with status {result.status}")
print(f"Team card: {card.name}")
print(f"Team endpoint: {card.url}")
print(f"Task ID: {result.task_id}")
print(f"Response: {result.content}")
# ---------------------------------------------------------------------------
# Run Team Server or Client Demo
# ---------------------------------------------------------------------------
if __name__ == "__main__":
if "--demo" in sys.argv:
asyncio.run(run_demo())
else:
agent_os.serve(app=app, port=7779)
Run the Example
1
Set up your virtual environment
uv venv --python 3.12
source .venv/bin/activate
uv venv --python 3.12
.venv\Scripts\activate
2
Install dependencies
uv pip install -U "agno[a2a,os]" openai
3
Export your OpenAI API key
export OPENAI_API_KEY="your_openai_api_key_here"
$Env:OPENAI_API_KEY="your_openai_api_key_here"
4
Run the example
Save the code above as
team.py, then run:python team.py