> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agno.com/llms.txt
> Use this file to discover all available pages before exploring further.

# The Context Company

> Send Agno traces to The Context Company for agent observability and production interaction analysis.

[The Context Company](https://www.thecontextcompany.com) is an AI agent observability and customer analytics platform that helps teams understand and improve agents in production. The Agno integration captures agent runs through OpenTelemetry and sends them to The Context Company.

## Setup

<Steps>
  <Step title="Install the integration">
    ```bash theme={null}
    uv pip install -U "contextcompany[agno]" agno openai yfinance
    ```
  </Step>

  <Step title="Export your API keys">
    <CodeGroup>
      ```bash Mac/Linux theme={null}
      export TCC_API_KEY="your_tcc_api_key_here"
      export OPENAI_API_KEY="your_openai_api_key_here"
      ```

      ```bash Windows theme={null}
      $Env:TCC_API_KEY="your_tcc_api_key_here"
      $Env:OPENAI_API_KEY="your_openai_api_key_here"
      ```
    </CodeGroup>
  </Step>

  <Step title="Initialize instrumentation">
    Call `instrument_agno()` before importing Agno.

    ```python theme={null}
    from contextcompany.agno import instrument_agno

    instrument_agno()

    from agno.agent import Agent
    from agno.models.openai import OpenAIChat
    from agno.tools.yfinance import YFinanceTools

    agent = Agent(
        name="Stock Price Agent",
        model=OpenAIChat(id="gpt-5.2"),
        tools=[YFinanceTools()],
    )

    agent.print_response("What is the current price of NVIDIA?")
    ```
  </Step>
</Steps>

## Captured data

The integration records:

* Agent runs and model calls
* Tool arguments, results, and errors
* Prompts and responses
* Token usage and latency
* Session, user, organization, and custom metadata

The Context Company analyzes production conversations and traces to surface recurring patterns and account-level insights. Teams can inspect the supporting runs and use that context to improve agent behavior.

## Run metadata

Use OpenInference's `using_attributes` context manager to associate a run with a session, user, organization, or named agent:

```python theme={null}
from openinference.instrumentation import using_attributes

with using_attributes(
    metadata={
        "tcc.sessionId": "session-123",
        "tcc.conversational": "true",
        "tcc.agent": "stock-price-agent",
        "tcc.userId": "user-123",
        "tcc.orgId": "org-456",
        "environment": "production",
    },
):
    agent.print_response("What is the current price of NVIDIA?")
```

## Developer resources

* [Complete Agno integration guide](https://docs.thecontextcompany.com/frameworks/agno)
* [Runnable Agno example](/examples/integrations/observability/the-context-company)
* [The Context Company Python SDK](https://pypi.org/project/contextcompany/)