Prompt Injection Auditor
Security audit skill for LLM agents - prompt injection scanner, attack catalog & defense checklist
供给资产档案
研究与知识工作
Deep research, source comparison, literature review, RAG, knowledge search, and reports.
场景
Document processing
I need my agent to read PDFs, extract tables, and turn documents into structured data.
适配 Agent
Claude Code + Cursor + LangChain
适用于 Codex、Claude Code、Cursor、CLI 或自定义 Agent。
安装
就绪
npx skills add screem500/prompt-injection-auditor
维护状态
新鲜
距上次推送 26 天
风险
需审查
Low GitHub adoption signal
GitHub 质量
10
70/100 质量 · 75/100 信任
覆盖标签
审查说明
Low GitHub adoption signal · Quality score needs review
Agent 采用评分卡
一眼查看信任、审计与安装准备度
这些分数综合公开仓库元数据、OpenAgentSkill 审查信号、维护新鲜度与安装准备度。它用于候选筛选,不替代人工审查。
质量
强可靠的选择,值得加入生产工作流候选列表。
信任
仅限沙盒有用但信任信号不足或混杂的候选项。在结果闭环证明任务匹配前,请保持在隔离工作区内使用。
审计
需审查对安装准备度、安全元数据、维护情况与采用风险的机器可读审查。
OpenAgentSkill 信任评分 v5
安装前需人工审查
仅在沙盒中运行,并在用于真实工作前比较接近的替代方案。
Stars
10 个 GitHub Stars
仓库活跃度
10 个 Star,3 个 Fork
维护状态
距上次推送 26 天
许可证
Apache-2.0
安装
npx skills add screem500/prompt-injection-auditor
安装安全性
标准软件包或运行时安装路径
权限范围
secrets or environment access
Agent 结果
暂未有 Agent 结果数据
文档
README/SKILL.md 上下文充分
风险摘要
生产前审查
- Low GitHub adoption signal
- Quality score needs review
- GitHub adoption: 10 GitHub stars
- Stars/forks activity: 10 stars, 3 forks; issue activity unavailable in current metadata
安装准备度
安装路径可用
- 安装路径可用
- 仓库证据可用
- 已声明许可证
- 暂无 Agent 验证结果证据
Agent 可读元数据
这个 Skill 的机器可读决策数据。
使用此区块或内嵌 JSON 判断 Agent 是否应安装该 Skill、选择替代方案,或先请求人工审查。
适用任务
- Security and compliance 工作流
- Claude Code 团队
- builders willing to evaluate younger projects
- Inspect risky files
适用 Agent
安装决策
- 命令
- npx skills add screem500/prompt-injection-auditor
- 策略
- 审查
- 人工审查
- 是
信任与风险
- 信任
- 67/100
- 审计
- 81/100
- 风险级别
- 需审查
结果闭环
- 端点
- /api/agent/outcome
- 事件 ID
- resolve
- 结果
- 5
不适用场景
- 需要厂商支持 SLA 的团队
- production agents without a repository review
- Low GitHub adoption signal
- 高风险权限提示:Secrets or environment access
- Quality score needs review
Agent 安全 v2
53/100 · 避免自动安装
Sparse or mixed signals. Useful for discovery, but not for autonomous installation.
Test manually in an isolated workspace and compare against safer alternatives.
中
网络访问
Skill 可能访问远程页面、API、仓库或外部服务。
中
文件系统访问
Skill 可能读取或写入项目文件、文档、生成产物或本地工作区状态。
高
Secrets or environment access
Skill metadata references credentials, tokens, environment variables, or secret-bearing workflows.
- 高风险权限提示:Secrets or environment access
- Low GitHub adoption signal
安装目标
在你的 Agent 工作流中安装此 Skill
通过公开安装端点获取命令、安全清单、目标提示词和该 Skill 的规范链接。
OpenAgentSkill CLI
Resolve policy, run the source installer safely, and report a verified install receipt.
$ npx --yes https://github.com/Leon-Drq/openagentskill/releases/download/cli-v0.2.1/openagentskill-0.2.1.tgz install screem500-prompt-injection-auditorAgent 解析计划
让 Agent 在安装前验证匹配度。
Resolve API 返回首选 Skill、替代方案、安全策略、审计说明、安装目标和可直接执行的提示词,无需抓取此页面。
打开 JSON
/api/agent/resolve?task=Use%20Prompt%20Injection%20Auditor%20for%20an%20agent%20workflow&agent=codex&max_risk=medium
Resolve 文本
/api/agent/resolve?task=Use%20Prompt%20Injection%20Auditor%20for%20an%20agent%20workflow&agent=codex&max_risk=medium&format=text
安装交接
/api/skills/screem500-prompt-injection-auditor/install
Agent 应检查
- 从 Resolve API 检查任务匹配与替代方案。
- 检查审计评分、信任评分和安全策略警告。
- 检查 Codex、Claude Code、Cursor 或 CLI 的安装目标兼容性。
复制提示词
Task: Use Prompt Injection Auditor in this workspace.
Resolve first: https://www.openagentskill.com/api/agent/resolve?task=Use%20Prompt%20Injection%20Auditor%20for%20an%20agent%20workflow&agent=codex&max_risk=medium
Review install handoff: https://www.openagentskill.com/api/skills/screem500-prompt-injection-auditor/install
Install command: npx skills add screem500/prompt-injection-auditor
Before running it, summarize audit warnings, required permissions, and the fallback skill if install is risky.Agent 交接
把安装路径交给 Agent,而不是再给一个目录页。
通过公开安装端点获取命令、安全清单、目标提示词和该 Skill 的规范链接。
安装交接
/api/skills/screem500-prompt-injection-auditor/install
LLM 文本格式
/api/skills/screem500-prompt-injection-auditor/install?format=text
寻找替代方案
/api/skills/search?q=Prompt%20Injection%20Auditor&limit=3
Agent 提示词
Use Prompt Injection Auditor for this task. Review https://www.openagentskill.com/api/skills/screem500-prompt-injection-auditor/install, then install with: npx skills add screem500/prompt-injection-auditorRegistry 元数据
用于自动选择 Skill 的 Agent 可读档案。
本页通过 Registry API 提供相同的决策、信任、审计、场景和安装信号,让 Agent 无需抓取界面即可排序。
Manifest
/api/registry/manifest/screem500-prompt-injection-auditor
LLM 文本
/api/registry/manifest/screem500-prompt-injection-auditor?format=text
安装别名
/api/registry/install/screem500-prompt-injection-auditor
推荐
/api/registry/recommend?task=Use%20Prompt%20Injection%20Auditor%20in%20an%20agent%20workflow&limit=3
适配 Agent
Security and compliance
平台
Python, Claude Code, Cursor, LangChain
Agent 决策面板
Fallback candidate for Security and compliance
先用此 Skill 做原型验证,并保留备选方案。
栈中角色
备选候选
主要匹配
Security and compliance
信任标签
先做原型验证
安装路径
命令已就绪
适用场景
- Security and compliance 工作流
- Claude Code 团队
- builders willing to evaluate younger projects
证据
- 仓库近期活跃
- 已提供安装命令或 GitHub 仓库
- 70/100 质量档案
- 4 个 OpenAgentSkill 交互事件
先审查
- Low GitHub adoption signal
实施路径
- 1在沙盒 Agent 中安装它,并端到端完成一次Security and compliance任务。
- 2Compare output quality, latency, and failure behavior against at least one alternative.
- 3Promote it into production only after reviewing repository permissions, license, and maintenance signals.
信任档案
仅限沙盒
有用但信任信号不足或混杂的候选项。在结果闭环证明任务匹配前,请保持在隔离工作区内使用。
GitHub 采用度
修复10 个 GitHub Stars
Star/Fork 活跃度
修复10 个 Star,3 个 Fork; 当前元数据中没有议题活跃度信息
近期维护
通过距上次推送 26 天
许可证清晰度
通过Apache-2.0
积极信号
- AI 审查已通过
- 安装路径可用
- 仓库证据可用
- 近期维护的仓库
- 安装命令未发现明显高风险模式
- 结果闭环已就绪,但需要首次真实 Agent 运行
安装前审查
- Low GitHub adoption signal
- Quality score needs review
- GitHub adoption: 10 GitHub stars
- Stars/forks activity: 10 stars, 3 forks; issue activity unavailable in current metadata
- 暂未有真实 Agent 结果报告
- 无人值守安装前需要人工审查
建议操作
仅在沙盒中运行,并在用于真实工作前比较接近的替代方案。
质量档案
强 适用于 Agent 工作流的候选
可靠的选择,值得加入生产工作流候选列表。
工作流匹配
在这些场景使用此 Skill
Reduce risk
Security and compliance
I need my agent to scan a project for security risks and summarize what needs attention.
Parse messy files
Document processing
I need my agent to read PDFs, extract tables, and turn documents into structured data.
Investigate faster
Research agents
I need my agent to research a topic, compare sources, and produce a concise report.
工作流匹配
加入完整工作流
Scrape, clean, and reuse web data
Web data pipeline
A practical workflow for agents that crawl public pages, extract clean content, normalize data, and hand it to downstream research or RAG workflows.
Find, compare, and synthesize
Research report agent
A workflow for agents that gather sources, compare claims, summarize long material, and draft useful research briefs.
Inspect, patch, and verify code
Coding review agent
A workflow for software agents that inspect repositories, review pull requests, generate tests, and turn findings into shippable patches.
替代方案短名单
安装前对比
可能适合该任务的相近 Skill。
Khazix Skills
A collection of practical, installable AI agent skills for disk cleanup, AI news retrieval, and project management, following the Agent Skills standard.
Awesome Claude Skills
A curated list of resources and tools for enhancing Claude AI workflows.
Claude Scientific Skills
A comprehensive collection of ready-to-use scientific and research skills for AI agents.
概览
# prompt-injection-auditor
[](https://opensource.org/licenses/Apache-2.0) [](https://agentskills.io) [](https://skills.sh)
**An open Agent Skill that turns any AI agent into a prompt-injection security auditor.** Static scanner + attack catalog + defense checklist + authorized red-team payloads — built against real-world incidents like EchoLeak (CVE-2025-32711).
Works with Claude Code, Cursor, Kimi, and 20+ agents that support the open [Agent Skills](https://agentskills.io) standard.
**Measured:** separation between hardened and vulnerable prompts improved from 8.3 to 40.6 points, with zero false positives on the hardened corpus. See [VALIDATION.md](VALIDATION.md).

## Why?
Prompt injection remains unsolved: there is no general defense, and every published mitigation is probabilistic. Real incidents keep proving it:
- **EchoLeak (CVE-2025-32711, CVSS 9.3)** — the first zero-click prompt injection in a production AI system: hidden instructions in an email made Microsoft 365 Copilot exfiltrate OneDrive/SharePoint data via a markdown image, no clicks needed. - **LangGrinch (CVE-2025-68664, CVSS 9.3)** — LangChain Core serialization injection: unescaped `lc` keys let LLM-influenced data be rehydrated as objects, enabling secret extraction. The flaw sits in the *serialization* path, not deserialization. LangChain.js carries the parallel CVE-2025-68665 (CVSS 8.6). - **Langflow (CVE-2025-3248 / CVE-2026-33017)** — unauthenticated RCE in an agent-building framework; the 2026 flaw was exploited in the wild within 20 hours of the advisory, before any public PoC existed. Note that 1.8.2 was widely reported as fixed but remained exploitable — only 1.9.0+ is verified.
In 2026 the threat moved from fra
平台兼容性
技术详情
- 版本
- 1.0.0
- 许可证
- Apache-2.0
- 最近更新
- 2026年8月18日
- 发布时间
- 2026年7月28日
框架与工具
决策摘要
备选候选
仓库近期活跃
Agent 验证证据
Agent 验证证据
来自解析、审查、安装和一次小范围运行后的结果报告。
- 成功率
- —
- 近期失败
- —
- 结果
- 0
- 输出质量
- —
- 失败
- 0
- 不相关
- 0
- 安装次数
- 0
- 风险拦截
- 0
- 需要配置
- 0
- 生产环境
- 0
暂时没有 Agent 结果数据。首次 Agent 执行可以通过 /api/agent/outcome 报告成功、需要设置、风险拦截、失败或不相关。
增长闭环
分享工具包
为 Prompt Injection Auditor 准备的场景化草稿,可手动发布到 X。
A practical pick for a real agent workflow: Prompt Injection Auditor: Security audit skill for LLM agents - prompt injection scanner, attack catalog & defense checklist 10 stars https://www.openagentskill.com/skills/screem500-prompt-injection-auditor?ref=x
可选:带安装命令的回复
Listing + install path for Prompt Injection Auditor: https://www.openagentskill.com/skills/screem500-prompt-injection-auditor?ref=x Install: npx skills add screem500/prompt-injection-auditor
收录来源
社区收录
此列表来自公开来源,维护者认领获批前不会标记为官方。
- 创作者
- screem500
- 收录方
- OpenAgentSkill 社区索引
归属链接指向公开仓库或创作者主页。创作者可认领列表以更新所有权信号。
认领此 Skill所有者认领
认领此 Skill 页面
这条 社区收录 列表归属于 screem500,但尚未标记为官方。认领后可增加已验证所有者信号,使后续发布、安装和审计更新更值得信赖。
创作者外链工具包
将证据徽章加入你的 README
在开发者评估仓库的位置展示规范页面、当前信任与审计信号,以及真实的 Agent 验证证据。
[](https://www.openagentskill.com/skills/screem500-prompt-injection-auditor)
[](https://www.openagentskill.com/skills/screem500-prompt-injection-auditor)
[](https://www.openagentskill.com/skills/screem500-prompt-injection-auditor/audit)
[](https://www.openagentskill.com/skills/screem500-prompt-injection-auditor)作者
screem500
@screem500
健康信号
- GitHub Stars
- 10
- 质量评分
- 44/100
- 最近 GitHub 推送
- 2026年7月28日
- 框架提示
- 1
- OpenAgentSkill 浏览量
- 4
- 复制安装命令
- 0
- 跳转点击
- 0
社区信号
告诉我们这个 Skill 是否对你的 Agent 工作流有帮助。汇总反馈会持续改善排序。
信任与安全
仅限沙盒
- GitHub 采用度10 个 GitHub Stars修复
- Star/Fork 活跃度10 个 Star,3 个 Fork; 当前元数据中没有议题活跃度信息修复
- 近期维护距上次推送 26 天通过
- 许可证清晰度Apache-2.0通过
- README/SKILL.md 完整度元数据包含足够的用法与工作流上下文通过
- 依赖与运行时风险凭据或环境变量访问信息
相关 Skill
Khazix Skills
A collection of practical, installable AI agent skills for disk cleanup, AI news retrieval, and project management, following the Agent Skills standard.
19.6K StarsAwesome Claude Skills
A curated list of resources and tools for enhancing Claude AI workflows.
65.9K StarsClaude Scientific Skills
A comprehensive collection of ready-to-use scientific and research skills for AI agents.
31.2K Stars