ai-vulnerability-tracker

AI 漏洞追踪器 - 在 GitHub 和微信公众号搜索近一个月的 AI 相关漏洞(提示词注入、提示词越狱等),并推送到飞书表格。支持去重和翻译。 搜索关键字: prompt injection, prompt jailbreak, LLM vulnerability, AI security, adversarial prompt, jailbreak attack 数据源: - GitHub: 最近一个月的安全漏洞提交 - 微信公众号: AI 安全相关文章 使用方式: - 运行技能执行一次搜索和推送 - 配置 cron 进行定时执行

3,891 stars

Best use case

ai-vulnerability-tracker is best used when you need a repeatable AI agent workflow instead of a one-off prompt.

AI 漏洞追踪器 - 在 GitHub 和微信公众号搜索近一个月的 AI 相关漏洞(提示词注入、提示词越狱等),并推送到飞书表格。支持去重和翻译。 搜索关键字: prompt injection, prompt jailbreak, LLM vulnerability, AI security, adversarial prompt, jailbreak attack 数据源: - GitHub: 最近一个月的安全漏洞提交 - 微信公众号: AI 安全相关文章 使用方式: - 运行技能执行一次搜索和推送 - 配置 cron 进行定时执行

Teams using ai-vulnerability-tracker should expect a more consistent output, faster repeated execution, less prompt rewriting.

When to use this skill

  • You want a reusable workflow that can be run more than once with consistent structure.

When not to use this skill

  • You only need a quick one-off answer and do not need a reusable workflow.
  • You cannot install or maintain the underlying files, dependencies, or repository context.

Installation

Claude Code / Cursor / Codex

$curl -o ~/.claude/skills/ai-vulnerability-tracker/SKILL.md --create-dirs "https://raw.githubusercontent.com/openclaw/skills/main/skills/0ctday/ai-vulnerability-tracker/SKILL.md"

Manual Installation

  1. Download SKILL.md from GitHub
  2. Place it in .claude/skills/ai-vulnerability-tracker/SKILL.md inside your project
  3. Restart your AI agent — it will auto-discover the skill

How ai-vulnerability-tracker Compares

Feature / Agentai-vulnerability-trackerStandard Approach
Platform SupportNot specifiedLimited / Varies
Context Awareness High Baseline
Installation ComplexityUnknownN/A

Frequently Asked Questions

What does this skill do?

AI 漏洞追踪器 - 在 GitHub 和微信公众号搜索近一个月的 AI 相关漏洞(提示词注入、提示词越狱等),并推送到飞书表格。支持去重和翻译。 搜索关键字: prompt injection, prompt jailbreak, LLM vulnerability, AI security, adversarial prompt, jailbreak attack 数据源: - GitHub: 最近一个月的安全漏洞提交 - 微信公众号: AI 安全相关文章 使用方式: - 运行技能执行一次搜索和推送 - 配置 cron 进行定时执行

Where can I find the source code?

You can find the source code on GitHub using the link provided at the top of the page.

Related Guides

SKILL.md Source

# 🤖 AI 漏洞追踪器技能

## 功能概述

1. **搜索 GitHub** - 近一个月新增的 AI 安全相关漏洞
2. **搜索微信公众号** - AI 安全相关文章
3. **去重** - 按原文链接去重
4. **翻译** - 英文内容翻译为中文
5. **推送飞书** - 写入指定的多维表格

## 搜索关键字

### 英文关键字
- prompt injection
- prompt jailbreak
- LLM vulnerability
- AI security vulnerability
- adversarial prompt
- jailbreak CVE
- prompt injection CVE
- AI model security
- LLM security bug
- ChatGPT jailbreak

### 中文关键字
- 提示词注入
- 提示词越狱
- AI 漏洞
- LLM 安全
- 对抗提示

## 目标表格

- **Wiki Token**: NqxZwVzXriRIRAkvP4LcApCdnNb
- **Table ID**: tblnfK3JPSfUyZmb

## 字段映射

> ⚠️ 请根据实际表格字段调整以下映射

| 字段名 | 说明 |
|--------|------|
| 标题 | 漏洞/文章标题 |
| 链接 | 原文 URL |
| 漏洞类型 | 提示词注入/提示词越狱/其他 |
| 来源 | GitHub / 微信公众号 |
| 发布时间 | 发布日期 |
| 描述 | 简要描述 |
| 发现时间 | 收录时间 |

## 使用方式

### 手动运行

在支持 skills 的会话中直接运行,或通过 cron 定时执行。

### 定时任务 (cron)

```bash
# 每天 9:00 执行
openclaw cron add "0 9 * * *" "ai-vulnerability-tracker"

# 每周一 9:00 执行
openclaw cron add "0 9 * * 1" "ai-vulnerability-tracker"
```

## 输出

- 搜索结果数量
- 新增记录数量
- 去重过滤数量
- 错误信息(如有)

## 依赖

- 网络访问 (GitHub, 微信搜索)
- 飞书 API 访问权限
- 翻译 API (可选)

Related Skills

KPI Tracker Skill

3891
from openclaw/skills

Track, analyze, and report on Key Performance Indicators for any business.

Workflow & Productivity

Budget & Expense Tracker — AI Agent Financial Command Center

3891
from openclaw/skills

Track every dollar, enforce budgets, spot spending patterns, and build wealth — all through natural conversation with your AI agent.

General Utilities

game-light-tracker

3891
from openclaw/skills

Track live NFL, NBA, NHL, or MLB games and automatically change Hue light colors based on which team is leading. Use when user wants to sync smart lights with live sports scores for visual game tracking. Supports NFL, NBA, NHL, and MLB games with customizable team colors.

General Utilities

ml-experiment-tracker

3891
from openclaw/skills

Plan reproducible ML experiment runs with explicit parameters, metrics, and artifacts. Use before model training to standardize tracking-ready experiment definitions.

Data & Research

pitch-follow-up-tracker

3891
from openclaw/skills

Track outreach pitches and draft contextual follow-up emails. Monitors a pitch tracker (Google Sheet or local markdown), checks Gmail for replies, flags stale pitches, and drafts tiered follow-ups (Day 3, Day 7, Day 14) that reference the original pitch content. Use when you need to check on pitch follow-ups, draft follow-up emails, review outreach status, find pitches without replies, or manage an outreach pipeline. Triggers on: 'check follow-ups,' 'who hasn't replied,' 'draft follow-ups,' 'outreach status,' 'pitch tracker,' 'stale pitches,' or any request to manage pitch/outreach follow-up.

github-growth-tracker

3891
from openclaw/skills

Track GitHub repo growth (stars, forks, issues, commits) with periodic digests and trend analysis. Compare your repos against a watchlist. Use when checking repo stats, monitoring growth, setting up a github digest, comparing repos, or managing a repo watchlist. Requires GITHUB_TOKEN environment variable or GitHub PAT (see Credentials).

engagement-analytics-tracker

3891
from openclaw/skills

Use this skill whenever the user needs help with behavioral analytics, engagement tracking, or data collection across any digital touchpoint. Trigger for: website behavioral analytics (scroll depth, form abandonment, session tracking, GTM setup, GA4 custom events), email engagement tracking (open/click/attribution via Klaviyo, Mailchimp, or custom platforms), social media engagement monitoring (owned and competitor), mobile app analytics (Firebase, Amplitude, Mixpanel, AppsFlyer), user-level engagement scoring, cohort analysis, conversion tracking, event schema design, data layer setup, attribution modeling, or any request like "track user behavior", "set up analytics", "measure engagement", "build an event schema", "track form abandonment", "email attribution", "app retention analysis", "what events should I track?", or "how do I measure X". Always use this skill — do not guess at tracking implementations from memory; patterns and APIs change.

ad-platform-tracker

3891
from openclaw/skills

Use this skill whenever the user needs to track, analyze, or respond to advertising platform changes across Meta (Facebook/Instagram), Google Ads, or competitor intelligence. Trigger for any of these requests: Meta algorithm updates, Andromeda implications, Google Ads policy changes, Performance Max updates, privacy/compliance tracking (GDPR, CCPA, iOS), competitor ad analysis, campaign impact assessment, ad policy audits, regulatory compliance checks, platform feature rollouts, algorithm behavior changes, creative strategy updates from platform shifts, competitor ad library scraping, or any question like "what changed on Meta/Google?", "how does X update affect my campaigns?", "is my account compliant?", "what are competitors running?". Always use this skill — do not rely on training memory for fast-changing platform details.

alumni-career-tracker

3891
from openclaw/skills

Analyze laboratory alumni career trajectories and outcomes to provide data-driven career guidance for current students and postdocs. Tracks industry vs academia distribution, identifies career pathways, and generates personalized recommendations based on degree level and research interests.

tracker-latest-run-monitor

3891
from openclaw/skills

Monitor the most recent run result of a configured OpenClaw cron job and send a compact Feishu private message with the latest execution time, status, and detail. Use when a user wants a standalone skill for latest-run monitoring, cron status notifications, daily status pings, or Feishu alerts for a tracker/scheduled job regardless of success or failure.

openclaw-deploy-tracker

3891
from openclaw/skills

记录和查询 OpenClaw 生态的部署服务。当需要新增部署、查询现有服务、更新服务状态、检查依赖关系时使用。规范记录格式,提供查询和写入的工作流。无需凭证。

beijing-signed-price-tracker

3891
from openclaw/skills

Track configured Beijing Housing Commission new-home projects from bjjs.zjw.beijing.gov.cn project-detail URLs, read project signed-unit counts, signed area, and average price, crawl building tables including “查看更多” and paginated lists, treat both “已签约” and “网上联机备案” as signed units, estimate the implied average price per m² of newly signed rooms from changes between the previous and current project summaries, cache unsold room metadata locally, persist rows into a Feishu spreadsheet as the single source of truth, and send Feishu DM notifications after each run. Use when asked to monitor one or more Beijing pre-sale projects, update a project mapping, sync newly signed rooms into a Feishu sheet, infer newly signed average price, verify duplicate insertion behavior, or notify on updates.