figure-it-out
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a procedural framework for task management and does not contain executable scripts or binary files.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes external tasks (representing an injection surface), it explicitly builds in defensive patterns such as falsifiable predicates, verification harnesses, and human-auditable logs to mitigate risks from untrusted data.
- [DATA_EXPOSURE]: The instructions reference internal paths for capability resolution (e.g.,
skills/poteto-mode/references/hosts/), but there are no attempts to access sensitive system files, environment variables, or hardcoded credentials. - [COMMAND_EXECUTION]: No shell commands, dynamic context injection (
!), or remote code execution patterns are present in the skill content.
Audit Metadata