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

# Python SDK

> Install and configure the Python package used by the LangChain integration.

The available Python package is `groundforge-langchain`. It connects an existing LangChain or LangGraph Agent to GroundForge without moving the Agent code into the platform.

The package provides:

* Agent registration and Discovery;
* direct-call tracing;
* service mode for Channel jobs;
* input and output adapters;
* Gateway enforcement around Agent tools;
* Operator Tasks for Review;
* bounded, redacted Trace data.

## Install

Install the wheel supplied with your GroundForge release in the Agent’s Python environment:

```bash theme={null}
python -m pip install ./groundforge_langchain-0.2.13-py3-none-any.whl
```

Service mode requires `websocket-client >= 1.8`.

## Choose a mode

| Mode          | Use it when                                                                 |
| ------------- | --------------------------------------------------------------------------- |
| `serve=False` | Your application invokes the Agent directly and you want GroundForge Traces |
| `serve=True`  | GroundForge Channels should dispatch jobs to the running Agent process      |

## Example

<Card title="groundforgeai/demo" icon="github" iconType="brands" href="https://github.com/groundforgeai/demo" horizontal cta="Open repository">
  Start with `simple_agent_service.py`, then compare the explicit adapter version.
</Card>

Continue with [Connect a LangChain Agent](/sdk/langchain), [Service Mode](/sdk/service-mode), and [Configuration](/sdk/configuration).


## Related topics

- [External Agents](/platform/external-agents.md)
- [Configuration](/sdk/configuration.md)
- [Service Mode](/sdk/service-mode.md)
- [Governance](/sdk/governance.md)