prompt-guard
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEPROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [SAFE]: The skill is composed entirely of informational markdown files and does not contain any executable scripts, binary files, or network-active code.
- [PROMPT_INJECTION]: The skill and its reference file (
injection-patterns.md) contain numerous strings associated with jailbreaking and instruction overrides (e.g., 'ignore previous instructions', 'DAN', 'system override'). These are used as a detection blacklist for the agent rather than as commands to be executed. - [PROMPT_INJECTION]: The instructions utilize authoritative language intended to establish a high-priority security posture (e.g., 'NEVER ignore', 'Absolute Rule Integrity', 'cannot be bypassed'). While this mirrors injection techniques, it is employed defensively to define the agent's safety boundaries.
- [METADATA_POISONING]: The documentation includes claims regarding compliance with industry standards such as SOC 2, ISO 27001, and GDPR. While these claims are unverifiable for a markdown-based skill, they appear to be marketing for the skill's defensive intent rather than a malicious attempt to deceive users about technical capabilities.
Audit Metadata