The Agent Skills Directory

[PROMPT_INJECTION]: No prompt injection patterns such as instructions to ignore safety guidelines or bypass restrictions were detected.
[DATA_EXFILTRATION]: No data exfiltration patterns or network operations to unknown domains were identified.
[CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or instructions to store credentials unsafely were found.
[OBFUSCATION]: No obfuscation techniques, including Base64 encoding, zero-width characters, or homoglyphs, were detected.
[REMOTE_CODE_EXECUTION]: No remote code execution or installation of untrusted packages was detected. The skill references potential scripts in its workflow description but does not provide or download any executable content.
[PRIVILEGE_ESCALATION]: No commands for acquiring elevated permissions, such as sudo or chmod, were present.
[EXTERNAL_DOWNLOADS]: The skill references public GitHub repositories for documentation and provenance tracking. These references target a well-known service and are documented for audit purposes.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-defined tasks for UX design. While this creates a potential surface for indirect injection, no malicious exploitation patterns were found, and no dangerous capabilities (such as arbitrary file writing or outbound network requests) are exposed by the skill itself.

ux-flow