Skip to main content
"""
Agent Input And Output Schemas
==============================

Demonstrates AgentOS agents that use input and output schemas.
"""

from typing import List

from agno.agent import Agent
from agno.db.sqlite import SqliteDb
from agno.models.openai import OpenAIChat
from agno.os import AgentOS
from agno.tools.hackernews import HackerNewsTools
from pydantic import BaseModel, Field

# ---------------------------------------------------------------------------
# Setup
# ---------------------------------------------------------------------------
input_schema_db = SqliteDb(
    session_table="agent_session",
    db_file="tmp/agent.db",
)

output_schema_db = SqliteDb(
    session_table="movie_agent_sessions",
    db_file="tmp/agent_output_schema.db",
)


class ResearchTopic(BaseModel):
    """Structured research topic with specific requirements."""

    topic: str
    focus_areas: List[str] = Field(description="Specific areas to focus on")
    target_audience: str = Field(description="Who this research is for")
    sources_required: int = Field(description="Number of sources needed", default=5)


class MovieScript(BaseModel):
    """Structured movie script output."""

    title: str = Field(..., description="Movie title")
    genre: str = Field(..., description="Movie genre")
    logline: str = Field(..., description="One-sentence summary")
    main_characters: List[str] = Field(..., description="Main character names")


# ---------------------------------------------------------------------------
# Create Agents
# ---------------------------------------------------------------------------
hackernews_agent = Agent(
    name="Hackernews Agent",
    model=OpenAIChat(id="gpt-5.2"),
    tools=[HackerNewsTools()],
    role="Extract key insights and content from Hackernews posts",
    input_schema=ResearchTopic,
    db=input_schema_db,
)

movie_agent = Agent(
    name="Movie Script Agent",
    id="movie-agent",
    model=OpenAIChat(id="gpt-5.2"),
    description="Creates structured outputs - default MovieScript format, but can be overridden",
    output_schema=MovieScript,
    markdown=False,
    db=output_schema_db,
)

# ---------------------------------------------------------------------------
# Create AgentOS
# ---------------------------------------------------------------------------
agent_os = AgentOS(
    id="agent-schemas-demo",
    agents=[hackernews_agent, movie_agent],
)
app = agent_os.get_app()

# ---------------------------------------------------------------------------
# Run
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    agent_os.serve(app="agent_schemas:app", port=7777, reload=True)

​
Run the Example

# Clone and setup repo
git clone https://github.com/agno-agi/agno.git
cd agno/cookbook/05_agent_os/schemas

# Create and activate virtual environment
./scripts/demo_setup.sh
source .venvs/demo/bin/activate

python agent_schemas.py