Secure your LLM applications with one line of code.
npm install @aibim/sdkimport OpenAI from 'openai';
import { wrap } from '@aibim/sdk';
const client = wrap(new OpenAI({ apiKey: 'sk-...' }), {
aibimUrl: 'https://your-aibim.example.com',
aibimApiKey: 'aibim-your-api-key',
});
// All LLM calls now route through AIBIM security proxy
const response = await client.chat.completions.create({
model: 'gpt-4o',
messages: [{ role: 'user', content: 'Hello!' }],
});wrap()/unwrap()for OpenAI and Anthropic clients- Direct prompt analysis (
AibimGuard) - Full management API (tenant, rules, alerts, data)
- Real-time WebSocket alerts
- Zero runtime dependencies (native fetch)
- Full TypeScript type safety
- Automatic retry with exponential backoff
- OpenAI
- Azure OpenAI
- Anthropic
- Any OpenAI-compatible API