prd-to-plan
Warn
Audited by Gen Agent Trust Hub on Apr 23, 2026
Risk Level: MEDIUMDATA_EXFILTRATIONNO_CODE
Full Analysis
- [DATA_EXFILTRATION]: Both
SKILL.mdandreferencesfiles contain relative path strings (../../../) pointing to locations outside the skill root directory. This is a directory traversal pattern which may allow unauthorized access to sensitive system files if the platform's file loader does not properly sanitize and restrict paths. - [NO_CODE]: The provided files contain no YAML frontmatter, markdown instructions, or tool definitions required for an AI agent skill to function, consisting solely of file path pointers.
Audit Metadata