social-sign-in-buttons
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and does not contain any executable code, shell commands, or automation scripts that could pose a security risk.
- [EXTERNAL_DOWNLOADS]: The skill references several official documentation sites and asset repositories (Apple, Google, Kakao, Naver, GitHub). These are well-known services and the skill correctly instructs the user to manually verify licenses and download assets rather than automating the process through the agent.
- [PROMPT_INJECTION]: No attempts to bypass safety filters or override system instructions were found. The language used is strictly for implementation guidance.
- [DATA_EXFILTRATION]: No sensitive file paths, environment variables, or hardcoded credentials are accessed or transmitted.
- [INDIRECT_PROMPT_INJECTION]: While the skill involves reviewing external documentation, it does not ingest untrusted dynamic data to perform sensitive operations. The risk of indirect injection via UI branding guidelines is negligible.
Audit Metadata