Security Scanner
Run automated security checks mapped to the OWASP API Top 10 and LLM security standards.
RaksHex includes an active runtime security scanner designed to detect vulnerabilities in your API routes and LLM prompts. By auditing inputs and responses in real-time, the scanner detects anomalies before they trigger critical data leaks or model compromises.
OWASP API Top 10 Coverage
Our scanner actively flags routes for:
- API1:2023 Broken Object Level Authorization (BOLA): Accessing records with unauthorized ID manipulation.
- API2:2023 Broken Authentication: Missing tokens, invalid validation, or signature bypass.
- API5:2023 Broken Function Level Authorization (BFLA): Standard users accessing administrative endpoints.
- API8:2023 Security Misconfiguration: CORS headers set to wildcard (
*), unencrypted payloads, or verbose stack traces in responses.
Prompt Injection Classifiers
The scanner uses a library of 87 distinct payloads representing indirect prompt injections, jailbreaks, and system prompt leakage attacks. We update this model weekly to capture new attack vectors targeting OpenAI, Anthropic, and Gemini models.