deepseek-pentest-ai-burp-extension

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The extension explicitly generates and automates exploit payloads (SQLi, RCE, SSRF, etc.), includes examples that can exfiltrate credentials, and sends captured request/response data to an external AI service with export features—a high-risk dual-use tool that can be abused for unauthorized data exfiltration and credential theft if misused or if redaction is incomplete.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The extension makes runtime calls to the DeepSeek API (e.g. https://api.deepseek.com/v1/chat/completions) to generate payloads/prompts that the extension then uses to drive attacks, so remote responses directly control generated instructions and are a required dependency.