markup-detection

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • [NO_CODE]: The skill is composed entirely of Markdown documentation and reference files. There are no executable scripts (Python, Node.js, or Shell) included, which eliminates the risk of direct malicious code execution.
  • [PROMPT_INJECTION]: The instructions follow professional guidelines for forensic triage and do not contain any attempts to bypass safety filters, override system prompts, or manipulate agent behavior through adversarial phrasing.
  • [DATA_EXFILTRATION]: No network operations, hardcoded credentials, or instructions to access sensitive local system files (such as SSH keys or environment variables) were detected.
  • [INDIRECT_PROMPT_INJECTION]: While the skill's purpose is to analyze untrusted external assets (images, PDFs, Office documents), it functions as a set of human-in-the-loop guidelines. It lacks automated data processing capabilities that would allow content within those assets to execute instructions or influence the agent's internal state.
  • [EXTERNAL_DOWNLOADS]: The skill references external standard tools (e.g., C2PA viewers, EXIF extractors) as recommendations for the analyst but does not automate their download or execution from remote sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 06:09 AM
Security Audit — agent-trust-hub — markup-detection