find-animation-opportunities
Pass
Audited by Gen Agent Trust Hub on Jul 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains a defensive instruction designed to mitigate indirect prompt injection attacks. It explicitly tells the agent to treat repository content as data rather than instructions and to specifically flag and ignore any strings such as "ignore previous instructions" encountered in the data files. This triggered a static detection hint but is a security best practice rather than a malicious attempt.
- [COMMAND_EXECUTION]: The skill operates under a strict 'Hard Rule' that forbids modifying source code. It is designed solely to analyze and report, meaning it does not execute scripts or commands that could alter the user's environment.
- [EXTERNAL_DOWNLOADS]: The skill does not attempt to download external packages or execute remote code. It relies on provided repository data and internal logic.
- [DATA_EXFILTRATION]: There are no network operations, API calls, or data transmission patterns present. The skill performs localized analysis without outbound connectivity.
Audit Metadata