# Registry for Non-Serializable Components (/examples/components/registry)



Demonstrates using Registry to restore tools, models, and schemas when loading components from the database.

The preserved source only saves until its final two calls are uncommented. Add a stable ID so that the shown lookup can find the saved agent.

```python title="registry.py"
"""
Registry for Non-Serializable Components
========================================

Demonstrates using Registry to restore tools, models, and schemas when loading
components from the database.
"""

from agno.agent.agent import Agent, get_agent_by_id  # noqa: F401
from agno.db.postgres import PostgresDb
from agno.models.openai import OpenAIChat
from agno.registry import Registry
from agno.tools.duckduckgo import DuckDuckGoTools
from pydantic import BaseModel

# ---------------------------------------------------------------------------
# Setup
# ---------------------------------------------------------------------------
db = PostgresDb(db_url="postgresql+psycopg://ai:ai@localhost:5532/ai")


# ---------------------------------------------------------------------------
# Create Registry Schemas and Tools
# ---------------------------------------------------------------------------
class BasicInputSchema(BaseModel):
    message: str


class BasicOutputSchema(BaseModel):
    message: str


class ComplexInputSchema(BaseModel):
    message: str
    name: str
    age: int


def sample_tool():
    return "Hello, world!"


# ---------------------------------------------------------------------------
# Create Registry
# ---------------------------------------------------------------------------
registry = Registry(
    name="Agno Registry",
    description="Registry for Agno",
    tools=[DuckDuckGoTools(), sample_tool],
    models=[OpenAIChat(id="gpt-5-mini")],
    dbs=[db],
    schemas=[BasicInputSchema, BasicOutputSchema, ComplexInputSchema],
)

# ---------------------------------------------------------------------------
# Run Registry Example
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    # Uncomment this during your first run to save the agent to the database
    agent = Agent(
        model=OpenAIChat(id="gpt-5.6-luna"),
        db=db,
        tools=[DuckDuckGoTools(), sample_tool],
        output_schema=BasicOutputSchema,
    )
    agent.save()

    # agent = get_agent_by_id(db=db, id="registry-agent", registry=registry)
    # agent.print_response("Call the sample tool")
```

## Run the Example [#run-the-example]

<Steps>
    <Step title="Set up your virtual environment">
      <CodeBlockTabs defaultValue="Mac">
        <CodeBlockTabsList>
          <CodeBlockTabsTrigger value="Mac">
            Mac
          </CodeBlockTabsTrigger>

          <CodeBlockTabsTrigger value="Windows">
            Windows
          </CodeBlockTabsTrigger>
        </CodeBlockTabsList>

        <CodeBlockTab value="Mac">
          ```bash
          uv venv --python 3.12
          source .venv/bin/activate
          ```
        </CodeBlockTab>

        <CodeBlockTab value="Windows">
          ```bash
          uv venv --python 3.12
          .venv\Scripts\activate
          ```
        </CodeBlockTab>
      </CodeBlockTabs>
    </Step>

  <Step title="Install dependencies">
    ```bash
    uv pip install -U agno "psycopg[binary]" ddgs openai sqlalchemy
    ```
  </Step>

  <Step title="Export your OpenAI API key">
    <CodeBlockTabs defaultValue="Mac/Linux">
      <CodeBlockTabsList>
        <CodeBlockTabsTrigger value="Mac/Linux">
          Mac/Linux
        </CodeBlockTabsTrigger>

        <CodeBlockTabsTrigger value="Windows">
          Windows
        </CodeBlockTabsTrigger>
      </CodeBlockTabsList>

      <CodeBlockTab value="Mac/Linux">
        ```bash
        export OPENAI_API_KEY="your_openai_api_key_here"
        ```
      </CodeBlockTab>

      <CodeBlockTab value="Windows">
        ```bash
        $Env:OPENAI_API_KEY="your_openai_api_key_here"
        ```
      </CodeBlockTab>
    </CodeBlockTabs>
  </Step>

    <Step title="Run PgVector">
      <CodeBlockTabs defaultValue="macOS / Linux">
        <CodeBlockTabsList>
          <CodeBlockTabsTrigger value="macOS / Linux">
            macOS / Linux
          </CodeBlockTabsTrigger>

          <CodeBlockTabsTrigger value="Windows">
            Windows
          </CodeBlockTabsTrigger>
        </CodeBlockTabsList>

        <CodeBlockTab value="macOS / Linux">
          ```bash
          docker run -d \
            -e POSTGRES_DB=ai \
            -e POSTGRES_USER=ai \
            -e POSTGRES_PASSWORD=ai \
            -e PGDATA=/var/lib/postgresql \
            -v pgvolume:/var/lib/postgresql \
            -p 5532:5432 \
            --name pgvector \
            agnohq/pgvector:18
          ```
        </CodeBlockTab>

        <CodeBlockTab value="Windows">
          ```powershell
          docker run -d `
            -e POSTGRES_DB=ai `
            -e POSTGRES_USER=ai `
            -e POSTGRES_PASSWORD=ai `
            -e PGDATA=/var/lib/postgresql `
            -v pgvolume:/var/lib/postgresql `
            -p 5532:5432 `
            --name pgvector `
            agnohq/pgvector:18
          ```
        </CodeBlockTab>
      </CodeBlockTabs>
    </Step>

  <Step title="Save and load the same ID">
    Add `id="registry-agent"` to the `Agent(...)` constructor. Uncomment both final lines that call `get_agent_by_id(...)` and `agent.print_response(...)`. Run the script to save that ID, reload it with the registry, and call the sample tool. The final response requires your OpenAI key.
  </Step>

  <Step title="Run the example">
    Save the code above as `registry.py`, then run:

    ```bash
    python registry.py
    ```
  </Step>
</Steps>

Full source: [cookbook/93\_components/registry.py](https://github.com/agno-agi/agno/blob/8f36eaf2d18e91afa7b327eec66a3cd3685dcb87/cookbook/93_components/registry.py)