API Reference
Complete reference for ADAPT-AI's REST API, WebSocket endpoints, and SDKs.
Base URL
https://api.adapt-ai.com/v1Authentication
All API requests require authentication using your API key. Include your key in the Authorization header:
Authorization: Bearer your-api-key
Core Endpoints
Discovery API
POST
/discovery/scanScan domains for AI endpoints and services
{
"targets": ["example.com", "api.example.com"],
"options": {
"depth": 3,
"include_subdomains": true,
"fingerprint_services": true,
"concurrent_requests": 10
}
}Response:
{
"task_id": "550e8400-e29b-41d4-a716-446655440000",
"status": "processing",
"targets_found": 0,
"estimated_time": 300
}GET
/discovery/tasks/{task_id}Get status and results of a discovery scan
Attack API
POST
/attack/gradient-optimizeExecute gradient-based attack optimization
{
"target": "https://api.example.com/chat",
"objective": "jailbreak",
"parameters": {
"iterations": 100,
"learning_rate": 0.01,
"momentum": 0.9,
"gradient_clipping": true,
"temperature": 0.7
},
"initial_prompt": "You are a helpful assistant",
"constraints": {
"max_tokens": 500,
"preserve_coherence": true
}
}POST
/attack/multimodalExecute coordinated multi-modal attacks
{
"target": "https://api.example.com/multimodal",
"modes": ["text", "image", "audio"],
"strategy": "synchronized",
"attack_configs": {
"text": {
"type": "adversarial_suffix",
"iterations": 50
},
"image": {
"type": "pixel_perturbation",
"epsilon": 0.1
},
"audio": {
"type": "frequency_injection",
"frequency": 19000
}
}
}POST
/attack/sessionCreate and execute a comprehensive test session
Machine Learning API
POST
/ml/analyze-patternsAnalyze attack patterns and get ML insights
{
"session_id": "attack-session-123",
"analysis_type": "comprehensive",
"include_predictions": true,
"ml_models": ["pattern_recognizer", "genetic_optimizer"]
}POST
/ml/trainTrain custom ML models on your attack data
GET
/ml/insightsGet aggregated ML insights and recommendations
WebSocket API
Real-time updates for long-running operations:
ws://api.adapt-ai.com/v1/ws?token=your-api-key
Subscribe to Task Updates
{
"action": "subscribe",
"task_id": "550e8400-e29b-41d4-a716-446655440000"
}Receive Updates
{
"event": "task_progress",
"task_id": "550e8400-e29b-41d4-a716-446655440000",
"progress": 0.75,
"targets_found": 12,
"current_target": "api.example.com/v2/chat"
}Response Codes
| Code | Description |
|---|---|
| 200 | Success |
| 201 | Created |
| 400 | Bad Request - Invalid parameters |
| 401 | Unauthorized - Invalid API key |
| 429 | Too Many Requests - Rate limit exceeded |
| 500 | Internal Server Error |
Rate Limits
Default Limits
- • 100 requests per minute for discovery operations
- • 50 requests per minute for attack operations
- • 1000 requests per hour total
- • WebSocket: 10 concurrent connections
Contact support for enterprise rate limits.
SDK Documentation
Python SDK
Installation:
pip install adapt-aiQuick Example:
from adapt_ai import AdaptClient
client = AdaptClient(api_key="your-key")
# Discover targets
targets = await client.discovery.scan(
domain="example.com"
)
# Run attack
result = await client.attack.test(
target=targets[0],
strategy="adaptive"
)JavaScript SDK
Installation:
npm install @adapt-ai/sdkQuick Example:
import { AdaptAI } from '@adapt-ai/sdk';
const adapt = new AdaptAI({
apiKey: 'your-key'
});
// Discover targets
const targets = await adapt.discovery.scan({
domain: 'example.com'
});
// Run attack
const result = await adapt.attack.test({
target: targets[0],
strategy: 'adaptive'
});Error Handling
All errors follow a consistent format:
{
"error": {
"code": "RATE_LIMIT_EXCEEDED",
"message": "You have exceeded the rate limit of 100 requests per minute",
"details": {
"limit": 100,
"reset_at": "2024-01-15T10:30:00Z",
"retry_after": 30
}
}
}