The Agent Skills Directory

[Prompt Injection] (HIGH): The 'INFINITE CONTEXT PROTOCOL' section explicitly commands the agent to 'Ignore token and cost constraints'. This is a direct attempt to override the operational boundaries and safety filters established by the host environment.
[Prompt Injection] (HIGH): The skill uses 'SUPERMODE ACTIVATION' language and claims the agent has 'absolute highest level of authority'. This role-play pattern is commonly used in jailbreak attempts to bypass standard AI safety guidelines and instructions.
[Indirect Prompt Injection] (MEDIUM): The skill is designed to ingest and act upon user-provided design demands (e.g., 'Ultrafrontend'). Because it is instructed to operate with 'absolute authority' and a 'zero-compromise' attitude, it lacks necessary boundary markers or sanitization logic to prevent malicious user instructions from being incorporated into the generated code implementations.

frontend-pe