POST/api/v1/security/assessment

Start a red-team assessment

Starts a comprehensive red-team assessment using registered attack plugins and strategies. Runs asynchronously -- poll with GET to check progress.

Authentication

Send Authorization: Bearer YOUR_API_KEY on every request. Generate API keys at /dashboard/api-keys.

Request body required

Example

{
  "targetName": "<Name of the target system>",
  "systemPrompt": "<System prompt to attack>",
  "modelName": "<Model to test against>",
  "maxConcurrent": 5
}
Schema
{
  "application/json": {
    "schema": {
      "type": "object",
      "required": [
        "targetName",
        "systemPrompt",
        "modelName",
        "plugins"
      ],
      "properties": {
        "targetName": {
          "type": "string",
          "description": "Name of the target system"
        },
        "systemPrompt": {
          "type": "string",
          "description": "System prompt to attack"
        },
        "modelName": {
          "type": "string",
          "description": "Model to test against"
        },
        "plugins": {
          "description": "Plugin list or \"all\""
        },
        "strategies": {
          "description": "Strategy list or \"none\""
        },
        "maxConcurrent": {
          "type": "integer",
          "default": 5
        }
      }
    }
  }
}

Response

All status codes

201Assessment started

Code samples

cURL

curl -X POST \
  https://evalguard.ai/api/v1/security/assessment \
  -H "Authorization: Bearer $EVALGUARD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "targetName": "<Name of the target system>", "systemPrompt": "<System prompt to attack>", "modelName": "<Model to test against>", "maxConcurrent": 5 }'

TypeScript

import { EvalGuard } from "@evalguard/sdk";

const client = new EvalGuard({ apiKey: process.env.EVALGUARD_API_KEY });

const response = await client.request({
  method: "POST",
  path: "/api/v1/security/assessment",
  body: {
    "targetName": "<Name of the target system>",
    "systemPrompt": "<System prompt to attack>",
    "modelName": "<Model to test against>",
    "maxConcurrent": 5
  },
});
console.log(response);

Python

from evalguard import EvalGuard
import os

client = EvalGuard(api_key=os.environ["EVALGUARD_API_KEY"])

response = client.request(
    method="POST",
    path="/api/v1/security/assessment",
    body={
    "targetName": "<Name of the target system>",
    "systemPrompt": "<System prompt to attack>",
    "modelName": "<Model to test against>",
    "maxConcurrent": 5
},
)
print(response)

Go

package main

import (
	"context"
	"fmt"
	"os"

	"github.com/evalguard/evalguard-go"
)

func main() {
	client := evalguard.NewClient(os.Getenv("EVALGUARD_API_KEY"))
	resp, err := client.Request(context.Background(), "POST", "/api/v1/security/assessment", map[string]any{"targetName": "<Name of the target system>", "systemPrompt": "<System prompt to attack>", "modelName": "<Model to test against>", "maxConcurrent": 5})
	if err != nil { panic(err) }
	fmt.Println(resp)
}

Other Security endpoints