security-review

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill's required output format asks the model to include "Evidence" code snippets from the files under review, which forces the LLM to reproduce any hardcoded secrets or tokens present in the provided code verbatim (creating an exfiltration risk), even though it doesn't explicitly request users' secrets otherwise.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). Most entries are legitimate documentation (OWASP, MITRE, vendor docs), but the set also contains clearly untrusted/malicious hosts and direct-download/script links (e.g., attacker.com, evil.com, attacker.com/script.sh, example.com/file.tar.gz), obfuscated/numeric IPs and metadata/internal endpoints (169.254.169.254, localhost/admin, internal-service) that are high-risk for malware distribution or SSRF-based credential/data exfiltration — treat any downloads or executables from those domains/URLs as suspicious.