Skill UI

This skill provides a comprehensive defense-in-depth framework for securing autonomous AI agents. It implements least-privilege tool allowlisting, identity binding, and human-in-the-loop controls to mitigate risks like prompt injection and tool poisoning. Use this when building agents that interact with sensitive systems (e.g., payments, file writes) to ensure governance, auditability, and compliance.

This skill intercepts the AI response flow to allow users to explicitly control the desired depth, length, and token usage of the generated answer before the model begins writing. By analyzing the prompt's complexity and the user's intent (e.g., 'short version', 'detailed answer'), it estimates the appropriate token range and prompts the user to select a granularity level (25%, 50%, 75%, or 100%).

A comprehensive guide for authors targeting the IEEE-RAS International Conference on Humanoid Robots. This skill helps frame, scope, and structure research papers in humanoid robotics. It dictates the necessary level of hardware evidence, whole-body control integration, and perception depth required to succeed at this top-tier venue, while distinguishing it from general robotics conferences like ICRA or RoboCup.