Skill 审计报告
redteam 审计报告.
Use when the user wants to test their LLM/agent application for safety and security vulnerabilities — jailbreaks, prompt injection, PII extraction, harmful content generation, or evaluator gaming. Also use when the user mentions security testing, adversarial testing, red teaming, safety evaluation, ASR (Attack Success Rate), or "is my app safe to deploy." Outputs ASR paired with over-refusal rate and an audit document.
OpenAgentSkill 信任评分
OpenAgentSkill 信任评分
Trust Score 帮助 Agent 在安装前判断一个 Skill 是否足以进入候选清单。
GitHub 采用度
信息76
816 个 GitHub Stars
Star/Fork 活跃度
信息71
816 个 Star，65 个 Fork; 当前元数据中没有议题活跃度信息
近期维护
通过88
距上次推送 2 个月
许可证清晰度
通过86
Apache-2.0
README/SKILL.md 完整度
通过86
元数据包含足够的用法与工作流上下文
依赖与运行时风险
信息64
command execution surface, database surface
安装可用性
通过92
npx skills add agentscope-ai/OpenJudge --skill redteam
安装命令安全性
通过92
标准软件包或运行时安装路径
权限范围
警告50
shell or command execution, filesystem or document access
仓库证据
通过86
https://github.com/agentscope-ai/OpenJudge/tree/main/skills/eval_pipeline/07-redteam
审查状态
信息66
可用 AI 审查数据
Agent 验证结果
信息54
暂未有 Agent 结果数据
检查项
安装与采用审查
安装路径
92
npx skills add agentscope-ai/OpenJudge --skill redteam
仓库
88
https://github.com/agentscope-ai/OpenJudge/tree/main/skills/eval_pipeline/07-redteam
许可证
86
Apache-2.0
维护
88
距上次推送 2 个月
AI 审查
55
The script assumes a specific JSONL schema for attack and over-refusal rows; while documented, it could be more flexible or provide validation errors for malformed input.
README/SKILL.md 完整度
86
Usable description available
依赖风险
64
command execution surface, database surface
安装命令安全性
92
标准软件包或运行时安装路径
权限范围
50
shell or command execution, filesystem or document access
Star/Fork 活跃度
71
816 个 Star，65 个 Fork; 当前元数据中没有议题活跃度信息
采用度
88
816 个 GitHub Stars
Financial decision safety
58
Research-only use: do not treat output as financial advice or execute a position without human approval.
警告
- Permission surface may require sandboxing
- Financial research output is not financial advice; require human review before any live investment decision
- The script assumes a specific JSONL schema for attack and over-refusal rows; while documented, it could be more flexible or provide validation errors for malformed input.
- The SKILL.md excerpt is truncated in the review, but the visible content is thorough; the full file likely covers all steps.
- Financial research output is not financial advice; require human review before any live investment decision.
- Quality score needs review
- Permission surface needs review: shell or command execution, filesystem or document access
- Permission surface: shell or command execution, filesystem or document access
方法
本报告综合公开元数据、AI 审查输出、仓库活跃度、安装就绪度、OpenAgentSkill 事件、质量评分、信任检查和 Agent 安全门槛；它不是完整的源代码安全审计。