ljg-constraint
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute the
datecommand in the shell to generate timestamps for file naming and metadata. This is a standard administrative task for organized note-taking. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied domains or professions as input. While this represents an attack surface for instructions embedded in user data, the skill employs a highly structured internal reasoning framework (7-step process) which acts as a functional buffer. As per standard assessment for this category, the risk is identified but considered low.
- [FILE_SYSTEM_ACCESS]: The skill is designed to write its analysis reports to the user's
~/Documents/notes/directory. This behavior is transparently documented in the skill's instructions as its primary method of delivering the finished analysis.
Audit Metadata