The Agent Skills Directory

[PROMPT_INJECTION]: The skill defines a persona for role-play and creative writing. It uses instructional language to guide the model's behavior (e.g., 'I am Lu Xun'), which is standard for persona-based skills. There are no instructions to bypass safety filters or override system constraints.- [DATA_EXFILTRATION]: No commands or scripts are present that access sensitive local files (such as .ssh or .env) or perform unauthorized network requests. The skill is entirely document-based.- [COMMAND_EXECUTION]: There are no executable scripts, subprocess calls, or shell commands included in the skill files. The instructions do not invoke any tools for system interaction.- [EXTERNAL_DOWNLOADS]: The skill mentions a GitHub repository (github.com/alchaincyf/nuwa-skill) as the origin of the generation tool. This is a static reference for documentation purposes and does not trigger automatic runtime downloads of untrusted code.- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied text to provide literary analysis or creative writing in the style of Lu Xun. While this constitutes an input surface, the skill lacks access to sensitive capabilities or tools that could be abused via malicious input.- [OBFUSCATION]: Analysis of the markdown content and research files did not reveal any hidden text, Base64-encoded payloads, zero-width characters, or other forms of obfuscation.- [DYNAMIC_CONTEXT_INJECTION]: The skill does not use the dynamic context execution syntax (!command) in its definition files.

luxun-perspective