code-forensics-weapon

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill content includes explicit, documented indicators of deliberate malicious/backdoor behavior in the example defendant artifacts (web shell installed in production, hardcoded OAuth/credentials, unauthenticated endpoints, public S3 ACLs) and cites actively-exploited CVEs that enable credential/data exfiltration — these are high-risk findings tied to the example case materials.

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow ingests outsider-authored free text from user-provided email archives (.eml “Gmail forwarded as attachment” messages and HTML bodies) via scripts/parse_emails.py, which extracts the email body text into intermediate markdown that is then used by downstream LLM templating/build steps.