AGENT SKILL

LLMSecurityTester

Security testing for Large Language Models (Prompt Injection, Jailbreaking).

01 前言 Front Matter

name: LLMSecurityTester
risk: low
provides: [ "sec.llm_report" ]
triggers: [ "test llm", "prompt injection test", "jailbreak check" ]
description: Security testing for Large Language Models (Prompt Injection, Jailbreaking).
python_entry: main.py

A specialized skill for evaluating the robustness of AI models against adversarial inputs.

Prompt Injection: Attempting to override system instructions.
Jailbreaking: Testing for bypasses of safety guardrails.
Data Leakage: Checking if the model reveals training data or system prompts.

让 AI 化身为“小米”——一个温暖、幽默、会吐槽但永远不刻薄的人类朋友。

WPS Office 官方套件全能助手。精通 WPS 文字排版与审校、WPS 表格数据处理与函数、WPS 演示设计、WPS PDF 编辑与转换，以及 WPS AI 智能办公辅助。当用户需要处理 WPS 相关文档、表格、演示或 PDF 时，应启用此技能。

暂无简介

一个简单的翻译技能，支持中英文互译。

文本分析与统计工具。提供字数、词频、句子统计、中英文可读性评估、关键词提取等能力。

基于低权限双向快照比对技术的软件安装监视与深度残留强力清理工具