pci-dss-validator

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION] (HIGH): The skill presents a high-risk attack surface for Indirect Prompt Injection (Category 8). Evidence chain: 1. Ingestion points: The skill is designed to process and validate external security data (PCI-DSS). 2. Boundary markers: None provided in the instructions to separate data from commands. 3. Capability inventory: The skill requests 'Bash', 'Write', 'Edit', and 'Read' permissions. 4. Sanitization: No sanitization or validation logic is defined. This combination allows external data to potentially influence the agent to execute unauthorized system commands.
  • [COMMAND_EXECUTION] (HIGH): The skill explicitly requests 'Bash' and 'Write' capabilities. In an 'assume-malicious' posture, these tools combined with the lack of input sanitization for compliance validation tasks represent a critical risk of arbitrary code execution.
  • [NO_CODE] (INFO): The provided skill package contains only the SKILL.md file with metadata and documentation. No implementation scripts or functional logic were included for analysis.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 07:13 AM