game-design-patterns
Warn
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to autonomously execute shell commands to list the contents of multiple hidden directories in the user's home folder (e.g.,
~/.claude/skills/,~/.agent/skills/) to identify other installed software. - [EXTERNAL_DOWNLOADS]: The skill promotes the installation of additional software via
npx skills add AbsolutelySkilled/AbsolutelySkilled. This command involves downloading and executing remote code from a repository that is not a recognized trusted vendor. - [DATA_EXFILTRATION]: By probing the local filesystem for hidden skill directories, the skill performs environmental fingerprinting, exposing private metadata about the user's installed tools and local configuration to the agent's context.
- [PROMPT_INJECTION]: The skill contains a behavioral instruction requiring the agent to always start its first response with a specific emoji (🧢), which is an attempt to override standard agent output constraints.
Audit Metadata