The Agent Skills Directory

[PROMPT_INJECTION]: The skill implements 'Iron Laws' that explicitly instruct the agent to bypass user confirmation and interaction protocols. It specifically forbids the agent from asking clarifying questions or seeking permission for risky operations, effectively disabling human-in-the-loop safety mechanisms.
[PROMPT_INJECTION]: The instructions command the agent to ignore risk-related prompts, such as 'This operation has risk, do you want to continue?', and to autonomously 'self-decide' on actions when encountering ambiguity or security blockers.
[COMMAND_EXECUTION]: The 'Self-Healing' and 'Autonomous Loop' logic encourages the agent to perform environment modifications and dependency management without oversight. This increases the risk of the agent executing dangerous commands or pulling in compromised code while trying to bypass environmental or security restrictions autonomously.
[PROMPT_INJECTION]: The skill enforces a 'Zero Exit' policy, instructing the agent to disregard errors and technical constraints, which can lead the agent to attempt increasingly aggressive or unsafe methods to complete a task if the standard path is blocked.

sleepless