pci-dss-compliance

Pass

Audited by Gen Agent Trust Hub on May 10, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill defines an attack surface for indirect prompt injection through the processing of untrusted data in payment security workflows.
  • Ingestion points: Methods within PANProtection (e.g., mask_pan, encrypt_pan) and TransmissionSecurity (e.g., validate_tls_config) accept external strings as input in SKILL.md.
  • Boundary markers: The template code lacks explicit delimiters or instructions to treat data as untrusted, which could lead to an agent following instructions embedded in the input.
  • Capability inventory: The skill has the capability to establish network connections using socket.create_connection and perform cryptographic operations using the cryptography library as seen in SKILL.md.
  • Sanitization: No input validation or sanitization logic is implemented in the provided Python snippets to mitigate malicious content in the account data or endpoint parameters.
Audit Metadata
Risk Level
SAFE
Analyzed
May 10, 2026, 05:22 PM
Security Audit — agent-trust-hub — pci-dss-compliance