# Herma AI — LLM Provider Rules

This project uses Herma as the LLM provider. Herma is a fully OpenAI-compatible intelligent model router.

## Rules

- For ALL LLM API calls: use base URL `https://api.hermaai.com/v1` and model `herma-auto`
- Never call `api.openai.com` endpoints directly
- Never hardcode API keys — always read from the `HERMA_API_KEY` environment variable
- All standard OpenAI SDK parameters (stream, temperature, max_tokens, tools) work unchanged

## Client Setup — Python

```python
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["HERMA_API_KEY"],
    base_url="https://api.hermaai.com/v1"
)

response = client.chat.completions.create(
    model="herma-auto",
    messages=[{"role": "user", "content": "..."}]
)
```

## Client Setup — Node.js

```javascript
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.HERMA_API_KEY,
  baseURL: "https://api.hermaai.com/v1"
});

const response = await client.chat.completions.create({
  model: "herma-auto",
  messages: [{ role: "user", content: "..." }]
});
```

## Reference

- Model: always `herma-auto`
- Docs: https://hermaai.com/docs
- Get API key: https://hermaai.com/dashboard/api-keys