harden
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of comprehensive documentation and guidelines for security research and software hardening. It does not contain any executable code, scripts, or malicious patterns.
- [INDIRECT_PROMPT_INJECTION]: The instructions specifically address Indirect Prompt Injection (IPI) risks. It mandates that the agent must never follow instructions found in code, logs, or external documents, treating them strictly as audit evidence rather than authority.
- [DATA_EXFILTRATION]: The 'Rules of Engagement' section explicitly prohibits the exfiltration, printing, or commitment of secrets, tokens, private keys, or PII. It provides safe patterns for reporting discovered credentials by noting only their location and type.
- [REMOTE_CODE_EXECUTION]: The skill includes guidance for identifying and remediating remote code execution vulnerabilities in target systems, specifically flagging insecure practices like 'curl-pipe-bash' in CI/CD pipelines.
- [PRIVILEGE_ESCALATION]: The methodology includes checks for broken access control and excessive agency, ensuring that the agent remains within authorized test boundaries and scopes.
Audit Metadata