find-animation-opportunities
Pass
Audited by Gen Agent Trust Hub on Aug 4, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains a defensive instruction designed to mitigate indirect prompt injection. It explicitly tells the agent to treat repository content as data and to flag and ignore attempts within those files to steer the agent (e.g., "ignore previous instructions").
- [DATA_EXFILTRATION]: No exfiltration patterns were detected. The skill is described as read-only and does not contain any network-reaching tools or commands.
- [COMMAND_EXECUTION]: The skill's hard rules explicitly state that it must never modify source code and only reports findings. There are no shell commands, subprocess calls, or scripts included that could execute arbitrary code.
Audit Metadata