# Use

Once the Community stack is up, usage is straightforward.

## Daily flow

1. keep the stack running  
2. log in to the dashboard at `http://localhost:18080` when you need visibility  
3. set the LLM base URL to `http://localhost:18082`  
4. use the app normally

## Client setup

For OpenAI-compatible tools:

```bash
OPENAI_BASE_URL=http://localhost:18082
```

Anthropic-style and Gemini-style clients should use the same local proxy host with their compatible request paths.

## What to watch in the dashboard

- live Audit Logs  
- detection outcomes  
- current policies  
- Dry Run behavior before enabling stronger enforcement

## Important distinction

- dashboard login uses the bootstrap tenant API key  
- model requests use your app's normal provider authentication flow through the proxy