The Agent Skills Directory

[SAFE]: The skill defines a set of instructions for the agent to conduct a technical interview in Chinese. It utilizes the AskUserQuestionTool as intended for legitimate user interaction. No code execution or network exfiltration patterns were detected.
[PROMPT_INJECTION]: The skill interpolates untrusted user data into instructions using the $ARGUMENTS variable. 1. Ingestion points: SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Agent interacts using AskUserQuestionTool. 4. Sanitization: Absent. This interpolation represents a minor surface for indirect prompt injection, but it is necessary for the skill's stated purpose of interviewing based on user requirements and does not pose a significant risk.

interview