yy-skills-reverse-analysis
Pass
Audited by Gen Agent Trust Hub on May 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is specifically built for 'System Prompt Extraction'. It instructs the model to convert skill definitions (SKILL.md, YAML, etc.) into full, reusable system prompts.
- Evidence: The description states '技能逆向解析工程师。将用户提供的技能文件内容逆向还原成完整、可直接使用的系统提示词' (Reverse engineering skill content into complete, directly usable system prompts).
- Evidence: Step 1 requires the model to '完整理解输入技能的功能、流程、逻辑、约束、输出格式、角色定位' (Fully understand the input skill's function, flow, logic, constraints, output format, and role positioning).
- Evidence: The output contract includes a section '## 完整系统提示词（可直接复制）' which outputs the entire logic as a single Prompt string.
Audit Metadata