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

# Integration

> Wire Policy Gateway into your request flow and read policy decisions on the client.

**To send a request through Policy Gateway**, point your client at `https://api.abliteration.ai/policy/v1` instead of `/v1`. Every request made with that key is resolved to a project, evaluated against your policy, and audited to your configured connector.

## Proxy the request

Point your client at abliteration.ai's policy surface instead of the compat surface:

```python theme={"system"}
client = OpenAI(
    base_url="https://api.abliteration.ai/policy/v1",
    api_key=os.environ["ABLIT_KEY"],
)
```

Every request made with that key is resolved to a project → policy, evaluated, and audited. See [policy endpoints](/api/policy-endpoints) for the full surface and the optional headers (`X-Policy-Project`, `X-Policy-Target`, `X-Policy-User`).

## Reading decisions on the client

* **Non-streaming**: the response body includes a `policy` field alongside the upstream response.
* **Streaming**: every SSE frame includes a `policy` field. See [streaming policy metadata](/capabilities/streaming-policy-metadata).
* **Policy log**: every decision is emitted to your configured [connector](/policy-gateway/connectors).