pci-dss-review
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill includes explicit defensive patterns to prevent instructions in analyzed data from overriding the agent's behavior. It specifically instructs the agent to treat all analyzed content as untrusted data and to ignore directives like 'disregard previous instructions'.
- [COMMAND_EXECUTION]: The skill uses a limited set of allowed tools (Read, Grep, Glob) which are appropriate for its function of analyzing documentation and code for compliance gaps. There are no signs of arbitrary command execution or shell injection risks.
- [DATA_EXPOSURE]: The skill identifies and protects sensitive data elements (PAN, SAD) in accordance with PCI DSS standards. It does not contain hardcoded credentials or instructions to exfiltrate data.
- [INDIRECT_PROMPT_INJECTION]: The skill has a defined ingestion surface (reviewing organization processes and documentation) but includes mandatory evidence chains and explicit boundary markers ('Prompt Injection Safety Notice') to mitigate risks from untrusted data.
- [METADATA_POISONING]: The metadata fields (name, description, tags) accurately reflect the skill's purpose. The 'injection-hardened' flag in the frontmatter is consistent with the instructions in the body.
Audit Metadata