pentest-expert

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The skill’s workflow runs public web reconnaissance/scanning tools (e.g., gobuster/feroxbuster/subfinder/amass/nikto/nuclei/sqlmap against http://target), which can fetch and ingest arbitrary outsider-authored page content and error responses into the agent’s runtime context as readable text.