attack-path-architect

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This skill explicitly provides actionable offensive techniques (SQLi→reverse shell, SSRF to cloud metadata to steal IAM credentials, cookie/session theft, CI/CD compromise, DNS takeover, etc.), enabling credential theft, remote code execution, and data exfiltration and therefore represents a high malicious risk despite being labeled for authorized testing.

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow can ingest OUTSIDER-authored free text when the user supplies only a raw domain, because it performs lightweight recon via recon-dominator and then feeds the resulting recon JSON (containing scraped/parsed web content and other third-party text) into the LLM context through the subsequent report-generation steps.