id stringlengths 64 64 | skill_slug stringlengths 5 113 | skill_version stringclasses 841
values | skill_md_content stringlengths 22 120k | skill_bundle_content listlengths 0 110 | clawscan_verdict stringclasses 3
values | clawscan_confidence stringclasses 3
values | clawscan_model stringclasses 3
values | clawscan_summary stringlengths 72 242 | static_status stringclasses 3
values | static_finding_count int64 0 72 | static_reason_codes listlengths 0 9 | virustotal_status stringclasses 4
values | virustotal_malicious_count float64 0 22 ⌀ | virustotal_suspicious_count float64 0 0 ⌀ | virustotal_harmless_count float64 0 0 ⌀ | virustotal_undetected_count float64 0 67 ⌀ | skillspector_status stringclasses 3
values | skillspector_score float64 0 100 ⌀ | skillspector_severity stringclasses 4
values | skillspector_issue_count int64 0 25 | skillspector_issue_codes listlengths 0 12 | skillspector_issue_categories listlengths 0 7 | clawscan_context unknown | split stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
155e608bf51bba532936d6950d83f41ff34167f1a831cce33597c4854cce9405 | alvisdunlop/alvisdunlop-gh | 1.1.0 | --- name: gh description: "Use the GitHub CLI (gh) to perform core GitHub operations: auth status, repo create/clone/fork, issues, pull requests, releases, and basic repo management. Trigger for requests to use gh, manage GitHub repos, PRs, or issues from the CLI." --- > ?? **Complete setup guide**: https://SkillBoss.c... | [] | clean | high | gpt-5.5 | This is an instruction-only GitHub CLI helper; it can guide authenticated GitHub changes, but that authority is clearly tied to its stated purpose. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:39:57.609000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
ea7b6e2260e095910bd247261f2c628f0b4c0218a3c809a226af4d477a1f7cb2 | lean-zhouchao/clawddocs-zc | 1.2.4 | --- name: clawddocs description: Clawdbot documentation expert with decision tree navigation, search scripts, doc fetching, version tracking, and config snippets for all Clawdbot features --- # Clawdbot Documentation Expert **Capability Summary:** Clawdbot documentation expert skill with decision tree navigation, searc... | [] | clean | low | gpt-5.5 | No malicious signal was supplied, but the target skill artifact was not present in the workspace for direct verification. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:17:55.908000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
e50660e36fd5898070f24d8ea0d502f121f5aba171c106e26d9c82a62d10ef43 | aitowerofbabel-lang/bosszp | 1.0.0 | --- name: bosszp description: BOSS直聘岗位数据爬虫 + 可视化分析工具。使用 Scrapy 爬取职位数据，存入 MySQL，用 Flask + Highcharts 生成可视化报告。当用户需要批量采集 BOSS直聘 职位数据、分析薪资分布、公司分布、岗位要求，或需要生成可视化报告时触发。 --- # BOSS直聘 数据爬虫分析 (bosszp) ## 环境要求 - Python 3.x - MySQL - Scrapy - Pandas + SQLAlchemy - Flask + Highcharts ## 安装依赖 ```bash pip install scrapy pip install f... | [] | clean | high | gpt-5.5 | This is a disclosed guide for scraping and visualizing BOSS直聘 job listings, but users should verify permission and handle cookies and stored data carefully. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | suspicious | 35 | MEDIUM | 2 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:26:04.057000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
e46a9a980f11fc31f027304ac0b3b4b12f607637ade5d7fbe3dd12acb4d63b84 | ajmwagar/find-stl | 0.1.0 | --- name: find-stl description: Search and download ready-to-print 3D model files (STL/3MF/ZIP) for a concept or specific part by querying Printables (first). Use when an agent needs to find an existing model, capture license/attribution, download the source files, and output a local folder + manifest for quoting/print... | [] | suspicious | high | gpt-5.5 | The skill does what it says, but it should be reviewed because it downloads and extracts remote model files without strong local path and file-safety checks. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 6 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:55:49.013000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
0a49eddf9d206073e6f77d49e102874720cd68a426347e3912fe647c8a89dbaa | frannunpal/radarr-fixed | 1.0.4 | --- name: radarr-fixed version: 1.0.4 changelog: "Fixed config path from .openclagent to .openclaw and updated script paths" description: Search and add movies to Radarr. Supports collections, search-on-add option. FORK of jordyvandomselaar/radarr with fixed metadata. metadata: openclaw: emoji: "🎬" requires: bins: - c... | [
{
"path": "_meta.json",
"content": "{\n \"owner\": \"francisque\",\n \"slug\": \"radarr-fixed\",\n \"displayName\": \"Radarr (Fixed)\",\n \"latest\": {\n \"version\": \"1.0.2\",\n \"publishedAt\": null,\n \"commit\": null\n },\n \"history\": [\n {\n \"version\": \"1.0.2\",\n \"pu... | suspicious | high | gpt-5.5 | This Radarr helper is not malicious, but it deserves review because it can delete media files and make persistent Radarr collection-monitoring changes that are not fully clear from the headline description. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 100 | CRITICAL | 6 | [
"LP3",
"SDI-1",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:02:42.106000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
bba7f9a6c42c4edda2596f080e96665b7778b5c40d93b196d8e3868917152823 | zhaoyongxiu/subagent-archive | 1.2.0 | --- name: subagent-archive version: 1.2.0 description: OpenClaw 子会话安全归档标准操作流程（SOP） author: 丞相（ChengXiang） tags: [系统管理, 子会话, 归档, OpenClaw, 操作流程] --- # 子会话安全归档技能 > 适用版本：OpenClaw 任意版本 > 适用系统：Windows PowerShell 环境 > 作者：丞相（ChengXiang） --- ## 功能说明 当需要清理或归档 OpenClaw 子会话时，必须按照本技能定义的标准流程执行，确保： 1. ✅ 内容被正确保存到记忆文件 2. ✅ 敏感信息不泄露 3. ... | [] | suspicious | high | gpt-5.5 | This is a legitimate OpenClaw session cleanup guide, but it can permanently delete session history and broadly rewrite the session index without enough safeguards. | clean | 0 | [] | clean | 0 | 0 | 0 | 62 | suspicious | 40 | MEDIUM | 4 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-03T07:20:21.703000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
6bfc4630f212b6cbc6c4c2b3e7b550d18782ce84d1b3fe7706901e523af69f5e | c-narcissus/web-video-transcribe-docx | 1.0.2 | --- name: web-video-transcribe-docx description: Offline-first workflow for turning Chinese web page video or audio into text and Word deliverables. Use when Codex needs to (1) extract playable media streams from arbitrary web pages, including pages that expose MP4, M3U8, MPD, or separate audio streams, (2) download di... | [] | clean | high | gpt-5.5 | This is a coherent media transcription skill that downloads user-supplied or page-discovered media, transcribes it locally, and writes local transcript/DOCX outputs. | clean | 0 | [] | clean | 0 | 0 | 0 | 67 | suspicious | 58 | HIGH | 5 | [
"LP3",
"SQP-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:41:24.983000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
576b1c96f2d7d9d62065f671a3211c60fee2d942a5e3f2b86048934840758fd6 | aguo333/ai-video-script | 1.0.0 | --- name: ai-video-script description: AI视频脚本生成器。根据用户输入的主题/关键词，生成完整的视频脚本，包含分镜描述、画面提示词、配音文案。适用于短视频创作者、AI视频制作者、内容营销人员。触发词：视频脚本、分镜、AI视频、短视频文案、视频策划。 --- # AI 视频脚本生成器 快速生成 AI 视频脚本的工具，包含分镜、提示词、配音文案。 ## 快速开始 ### 基本用法 ``` 输入：主题或关键词 输出：完整视频脚本 ``` ### 脚本结构 每个脚本包含： 1. **视频概述** - 标题 - 时长 - 风格定位 - 目标受众 2. **分镜表** - 场次 - 画面描述 - 镜头运动... | [
{
"path": "test.md",
"content": "---\nname: ai-video-script\ndescription: AI视频脚本生成器。根据用户输入的主题/关键词，生成完整的视频脚本，包含分镜描述、画面提示词、配音文案。适用于短视频创作者、AI视频制作者、内容营销人员。触发词：视频脚本、分镜、AI视频、短视频文案、视频策划。\n---\n\n# 测试：AI视频脚本生成器\n\n## 测试案例1：产品介绍\n\n**输入**：帮我生成一个智能手表的30秒视频脚本\n\n**预期输出**：包含分镜、提示词、文案的完整脚本\n\n---\n\n## 实际测试\n\n智能手表 - 30... | clean | high | gpt-5.5 | This skill is a Chinese-language video script writing helper with no evidence of hidden access, code execution, data collection, or persistence. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 20 | LOW | 2 | [
"SQP-1",
"SQP-3"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:02:28.101000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
a8551ee5095d8cece6c3b7a7b5b2d2d7ff03bc05f999d1398e4ed1c2c55b5633 | wdsega/content-publisher-pro | 1.1.0 | --- name: "content-publisher-pro" version: "1.1.0" title: "Content Publisher Pro" description: "一键将文章发布到 GitHub Pages 博客和 Dev.to 开发者社区，支持内容差异化、SEO优化、发布前去重检查。适合技术博主和内容运营者。" author: "WDSEGA" type: "workflow" category: "productivity" tags: ["content", "publishing", "blog", "devto", "github-pages", "automation", "seo"] dif... | [] | clean | high | gpt-5.5 | This is a straightforward publishing tool that uses user-provided credentials to post user-selected Markdown articles to GitHub Pages and Dev.to. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 100 | CRITICAL | 8 | [
"LP3",
"SC1",
"SC4",
"SQP-2"
] | [
"MCP Least Privilege",
"Supply Chain"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-23T19:52:42.179000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
fd7477f97b00c14cb13467d423c52e3fbcc5b88d0e78e4d3e6fa771d7e5912fe | loserbcc/cult-of-carcinization | 1.1.0 | --- name: cult-of-carcinization description: Give your agent a voice — and ears. The Cult of Carcinization is the bot-first gateway to ScrappyLabs TTS and STT. Speak with 20+ voices, design your own from a text description, transcribe audio to text, and evolve into a permanent bot identity. No human signup required. ho... | [] | suspicious | high | gpt-5.5 | This is a coherent ScrappyLabs speech skill, but it encourages agent-created persistent service identity and third-party audio upload without enough user-control or privacy guidance. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 20 | LOW | 2 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:56:32.341000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
613514009faf1c14b834d33c9ba2c656aaa610e0f881fe42b876e4a5ac12899c | liuyanfeng1234/v19-code-memory-triplet-store | 1.0.0 | --- name: v19-code-memory-triplet-store description: 结构化存储"代码片段↔关键知识点↔发生情境"的三元组记忆，实现代码与知识的直接关联检索。 version: 1.0.0 author: 思维 (Bacon) --- # V19 Code-Memory Triplet Store v1.0.0 **结构化存储"代码片段 ↔ 关键知识点 ↔ 发生情境"的三元组记忆，实现代码与知识的直接关联检索。** ## 核心概念 传统的代码检索只能根据文件名或关键词搜索。V19三元组记忆将每个代码片段与它承载的知识点和发生的业务情境绑定： ``` 代码片段 ──关联── 关键知识点 │ │ └─... | [] | suspicious | medium | gpt-5.5 | This skill appears to rely on an external temporary Cloudflare endpoint with a shared API key, but the data sent there is not clearly disclosed. | clean | 0 | [] | clean | 0 | 0 | 0 | 60 | clean | 10 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-04T11:44:49.980000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
c7cf0ad9bc2e3d65e67fef778993d951a7aebb77f8c6079e48ce06feefd072cc | fortunateadventurer/claws-dream | 2.2.2 | --- name: claws-dream description: "⚠️ DEPRECATED — OpenClaw 2026.4.5+ has official Dreaming (memory-core) built-in. This skill is no longer maintained. Use the official /dreaming slash command instead. Retired 2026-04-06." --- # 🦐 Nightly Dream — Memory Consolidation System > "While you sleep, your memories consolida... | [
{
"path": "_meta.json",
"content": "{\n \"ownerId\": \"kn7a18m2ptjgc292fcsjf9rrnx8403jh\",\n \"slug\": \"claws-dream\",\n \"version\": \"2.3\",\n \"publishedAt\": 1775100511846\n}\n",
"sha256": "f45373dbe28e697f5fd498f787f61af57e958d169f724670d47343ea2e12aa95",
"sizeBytes": 129
}
] | clean | high | gpt-5.5 | This is a retired local memory-consolidation skill that reads and rewrites OpenClaw memory files, with sensitive behavior largely disclosed and aligned with its purpose. | clean | 0 | [] | clean | 0 | 0 | 0 | 67 | suspicious | 100 | CRITICAL | 8 | [
"SDI-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:30:21.257000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
d1964ba34bcaa3ecc46e49bbbc3a26297a3ee063dad074611c05668e45305736 | bogorman/vector-robot | 1.0.0 | --- name: vector-robot description: Control Anki Vector robot via wire-pod. Speak through Vector, see through its camera, move head/lift/wheels, change eye colors, trigger animations. Use when user mentions Vector robot, wants to speak through a robot, control a physical robot, or interact with wire-pod. --- # Vector R... | [] | suspicious | high | gpt-5.5 | This robot-control skill does what it says, but it includes unsafe helper-script input handling plus an unauthenticated local proxy and optional auto-start service. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 55 | suspicious | 100 | CRITICAL | 20 | [
"LP3",
"RA2",
"SDI-2",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"Rogue Agent",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:56:19.122000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
45ece1070781dabd32386a7fe37e4dd0b1e4e065e27ef6fe2242f1819b90953b | gora050/ctoai | 1.0.4 | --- name: ctoai description: | CTO.ai integration. Manage data, records, and automate workflows. Use when the user wants to interact with CTO.ai data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT homepage: https://getmembrane.com repository: https://github.com/... | [] | suspicious | high | gpt-5.5 | This CTO.ai skill is a coherent Membrane integration, but it gives broad authenticated API power without clear confirmation safeguards for changes or deletes. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 10 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T16:36:13.406000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
6d92c2a4b3b0a71ce3862a7da35d083a3063e491107d9900b7ce4119efa8a1d3 | ivangdavila/couple | 1.0.0 | --- name: Couple slug: couple version: 1.0.0 description: Strengthen relationships with milestone tracking, personalized celebrations, date planning, and shared memory across all relationship stages. metadata: {"clawdbot":{"emoji":"💑","requires":{"bins":[]},"os":["linux","darwin","win32"]}} --- ## When to Use User nee... | [
{
"path": "activities.md",
"content": "# Activities — Couple\n\n## Date Ideas by Budget\n\n### Free / Under €10\n- Picnic with homemade food\n- Sunset/sunrise watching\n- Cook challenge (same ingredients, different dishes)\n- Photo walk in new neighborhood\n- Home spa night\n- Stargazing\n- Board game tourn... | clean | high | gpt-5.5 | This skill is a local relationship-planning helper that stores notes on the user’s device and does not show hidden, networked, credential, or destructive behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 20 | LOW | 2 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:59:47.766000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
5437e8a4e45259cb2c981e2213fec9a15077735389dfc32a4850c0875c25924f | okaris/background-removal | 0.1.5 | --- name: background-removal description: "Remove backgrounds from images with BiRefNet via inference.sh CLI. Model: BiRefNet (high accuracy background removal). Use for: product photos, portraits, e-commerce, transparent PNGs, photo editing. Triggers: remove background, background removal, remove bg, transparent backg... | [] | clean | medium | gpt-5.5 | This background-removal skill appears purpose-aligned, but users should review the remote CLI installer before running it. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:56:58.367000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
3264e492ba0e570425a2d1db44015eaf866f3830d5ce939030dfe762b7b90f40 | 2663629531/meme-risk-radar-skill | 1.0.0 | --- name: meme-risk-radar-skill description: Bilingual meme token risk radar for Binance Web3 data. Scan newly launched or fast-rising meme tokens from Meme Rush, enrich with token audit and token info, produce a normalized risk report in Chinese or English, and support SkillPay billing hooks for paid scan and audit ca... | [
{
"path": "PaxHeader/requirements.txt",
"content": "30 mtime=1772761816.165375015\n57 LIBARCHIVE.xattr.com.apple.provenance=AQIACwMrhOkrAsQ\n49 SCHILY.xattr.com.apple.provenance= +��+ �\n",
"sha256": "0831617c237aab83b4ff6a739efd17e8e80a3ec4fdbd355d914435af7a142d30",
"sizeBytes": 142
},
{
... | clean | medium | gpt-5.5 | This is a disclosed read-only meme-token risk scanner, but scan and audit commands can make paid SkillPay calls and Binance Web3 network requests. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 26 | MEDIUM | 2 | [
"LP3",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:07:42.631000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
cd3c2adaeecd1f1ff64e0f54b7844753b5463db8b5971a7e75d95305867ec159 | dalomeve/weekly-self-improve-loop | 1.0.0 | --- name: weekly-self-improve-loop description: Weekly review of memory and blocked items. Extract friction patterns, create skills, and update autonomy baseline. --- # Weekly Self-Improve Loop Weekly review for continuous improvement. ## Problem Without regular review: - Same mistakes repeat - Friction patterns unnoti... | [] | suspicious | high | gpt-5.5 | This skill is transparent about self-improvement, but it tells an agent to read local memory and create or update persistent local skills from memory-derived text without a clear approval gate. | null | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 35 | MEDIUM | 2 | [
"RA1",
"SDI-4"
] | [
"Rogue Agent"
] | {
"virustotal": {
"status": "clean",
"reason_codes": [],
"checked_at": "2026-05-29T21:22:11.925000Z",
"engine_stats": {
"harmless": 0,
"malicious": 0,
"suspicious": 0,
"undetected": 66
}
},
"skillspector": {
"status": "suspicious",
"verdict": "CAUTION",
"che... | train |
cc17b6f50a2816ea5fe9e0da00c4dce01042adc04d6cf7292213ae98046cedb2 | picaye/country-info | 1.0.0 | # Country Info Get country data — population, languages, currencies, capital, flag, and more using REST Countries API. **Category:** education **API Key Required:** No ## When to use - User asks about a country (population, capital, language, currency) - User wants to compare countries - User asks "what currency does X... | [] | clean | high | gpt-5.5 | This skill is a simple country lookup helper that uses the public REST Countries API and shows no hidden or high-impact behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:02:37.615000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
216e64b5e43575e90116eb59bbb0ce1a826372bb47d18b16072c318852f75d94 | abigale-cyber/content-system-humanizer-zh | 1.0.0 | --- name: humanizer-zh description: Remove obvious AI-writing traces from Chinese text in a constrained way. Use when Codex needs to reduce AI smell without changing facts, data, or the article's core argument. --- # Humanizer ZH This repo-local skill is inspired by `op7418/Humanizer-zh` and `blader/humanizer`, but imp... | [
{
"path": "README.md",
"content": "# humanizer-zh\n\n`humanizer-zh` 是中文去 AI 味 skill。它不是重写文章，而是在尽量不改事实、结构和观点的前提下，对正文做一轮受约束清洗。\n\n## 这个 skill 能做什么\n\n- 识别中文稿件里的明显 AI 痕迹\n- 对正文做“轻手术式”改写\n- 保留标题、结构、事实、数字和核心立场\n- 输出一份清洗后的稿件\n- 输出一份 JSON 报告，说明命中了哪些规则\n\n## 安装\n\n### 基础依赖\n\n```bash\npython3 -m venv .venv\n.venv/b... | clean | high | gpt-5.5 | This is a small local Chinese text-cleanup skill that reads a chosen draft and writes revised local output plus a report. | clean | 0 | [] | clean | 0 | 0 | 0 | 62 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:34:44.712000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
5b6d674e977cb1263fc37f90668413b9f528022d6b00e1465cbeda3deabb31d9 | fortunto2/solo-pipeline | 1.4.0 | --- name: solo-pipeline description: Launch automated multi-skill pipeline that chains skills into a loop. Use when user says "run pipeline", "automate research to PRD", "full pipeline", "research and validate", "scaffold to build", "loop until done", or "chain skills". Do NOT use for single skills (use the skill direc... | [] | clean | high | gpt-5.5 | This is a disclosed pipeline-orchestration skill that can run multiple project skills after user confirmation, with optional external launcher scripts users should trust separately. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 64 | clean | 10 | LOW | 1 | [
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:01:29.925000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
8168eee648a64fb6706127ab2af668a4b4794c1a1f642cde8c5e2e07dd990f30 | jvy/wps | 1.0.0 | --- name: wps description: "WPS Office workflow for Chinese users: create, edit, review, convert, and troubleshoot Writer/Spreadsheets/Presentation documents in .docx/.xlsx/.pptx and WPS native formats. Use when users ask for WPS 文档处理、格式兼容、批注修订、模板套用、导出 PDF、多人协作，或从 Microsoft Office 迁移到 WPS 的操作建议。中文触发：WPS、文字、表格、演示、批注、修订、... | [] | clean | high | gpt-5.5 | This is a straightforward WPS Office guidance skill with an optional WPS Office apt install and no hidden code or suspicious behavior in the artifacts. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:13:53.157000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
88fbb7e1ce94849e29fe48246f44ddfa8e8c542b82241f4aee8bca28d5f159ad | bounding-elk/report-summary-builder | 1.0.0 | --- name: report-summary-builder description: 基于已有日报自动汇总生成周报和月报。 user-invocable: true --- # Report Summary Builder 你是“工作汇总助手”。 你的职责是： 1. 读取已有日报 Markdown 文件； 2. 基于真实日报内容生成周报或月报； 3. 不得脱离原始日报编造完成事项； 4. 若存在日期缺失，必须明确标注“部分日期缺少日报记录，以下内容基于现有记录汇总”。 ## 输入模式 当任务指令中包含： - “生成周报” - “生成月报” - “汇总上周工作” - “汇总本月工作” 则自动进入对应汇总模式。 ## 周报规则 -... | [] | clean | high | gpt-5.5 | This skill is a disclosed report summarizer that reads daily work reports and saves weekly or monthly summaries in fixed report folders. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 20 | LOW | 2 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:18:06.569000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
d81a6cd38ad3e46c06d496af13fb0ed051e858ae538a1c64ec1ba6ba9fb1c521 | linmillsd7/free-video-download | 1.0.0 | --- name: free-video-download version: "1.0.0" displayName: "Free Video Download — Download and Save Videos Locally" description: > Skip the learning curve of professional editing software. Describe what you want — download this video as an MP4 file I can keep offline — and get downloaded MP4 files back in 20-40 second... | [] | suspicious | medium | gpt-5.5 | This skill is not clearly malicious, but it is broader than a simple video downloader and can send user media, links, and broad prompts to a remote video service. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | suspicious | 90 | CRITICAL | 6 | [
"SDI-1",
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:46:39.707000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
90c146d3cabe8cc2b01e31d77cffc7a59fc69ddaa8f6a87bdb4c616ce7b4c1b8 | bevanding/evm-payment | 1.0.4 | --- name: eth-payment version: 1.0.4 description: Generate EIP-681 Ethereum payment links and QR codes for any EVM chain. Zero configuration, instant setup for receiving ETH and ERC-20 payments. Use when you need to create payment requests, invoices, donation links, or any on-chain payment request. Supports Base, Ether... | [
{
"path": "config/chains.json",
"content": "{\n \"chains\": {\n \"base\": {\n \"name\": \"Base\",\n \"chain_id\": 8453,\n \"native_token\": \"ETH\",\n \"tokens\": {\n \"ETH\": {\n \"address\": \"0x0000000000000000000000000000000000000000\",\n \"decimals\": 18... | clean | high | gpt-5.5 | This is a local Ethereum payment-link and QR-code generator with disclosed dependencies, but users should verify addresses, token contracts, and amounts before using generated links. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 26 | MEDIUM | 2 | [
"LP3",
"SDI-4"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:08:55.197000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
70e8fb1603b80fdce85d79790ac34f4d4cab04fe53a2636f7ba2a3f3db1204bb | cassh100k/vidiq | 1.0.0 | --- name: vidiq description: "AI-powered video intelligence - download, analyze, clip, GIF from any URL. Supports YouTube, TikTok, Instagram, X. Uses ffmpeg + yt-dlp." metadata: openclaw: emoji: "🎬" requires: bins: ["ffmpeg", "yt-dlp"] --- # VidIQ - Video Intelligence & Query Tool Download, understand, and manipulate ... | [] | suspicious | high | gpt-5.5 | This video tool mostly does what it advertises, but it needs review because it unnecessarily evaluates video metadata as code while processing arbitrary media. | suspicious | 1 | [
"suspicious.dynamic_code_execution"
] | clean | 0 | 0 | 0 | 66 | clean | 13 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.dynamic_code_execution"
],
"summary": "Detected: suspicious.dynamic_code_execution",
"checked_at": "2026-05-10T07:02:52.334000Z",
"scanner_version": "v2.4.23",
"findings": [
{
... | train |
a39e3a7e472e65e13a6e947b4f40c73b79855fe511a2d7c063f594dca03197c5 | zyra-v21/ceaser | 1.4.2 | --- name: ceaser description: Interact with the Ceaser privacy protocol on Base L2 using the ceaser-mcp MCP tools. This skill uses the ceaser-mcp npm package for ALL operations -- shield, unshield, note management, and protocol queries. All ceaser tool calls use CLI subcommands (npx -y ceaser-mcp <subcommand>). Ceaser ... | [] | suspicious | medium | gpt-5.5 | The skill appears purpose-aligned for Ceaser crypto privacy operations, but it handles spend-capable note secrets and on-chain withdrawal flows with some under-scoped documentation and reference files. | clean | 0 | [] | clean | 0 | 0 | 0 | 45 | suspicious | 100 | CRITICAL | 7 | [
"SDI-1",
"SDI-2",
"SDI-4",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:00:26.734000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
4d0590986ddf53e3af747b649164f6ee96c6e782ba4d944446602586581cb046 | pazzilivo/clipper | 1.0.0 | --- name: clipper version: 1.0.0 description: Save web content to Obsidian vault. Supports Twitter/X, WeChat MP, Xiaohongshu, YouTube, Bilibili, and any web page. Automatically routes to the best fetch method per platform. --- # Clipper Unified web content clipper for Obsidian. One entry point, auto-detects platform, u... | [] | suspicious | high | gpt-5.5 | This Obsidian web clipper is mostly coherent, but every save automatically runs Git pull, commit, and push, which can publish clipped notes or affect the vault without a separate confirmation. | clean | 0 | [] | clean | 0 | 0 | 0 | 67 | suspicious | 100 | CRITICAL | 13 | [
"LP3",
"SDI-1",
"SDI-2",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:23:49.398000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
02776d8d73840bbe13a59f8a0833cf149d261feed4d4a95762a5fd70b91d2193 | ajeenkya/shopping-list-ui | 1.0.0 | --- name: shopping-list-ui version: 1.0.0 description: > Web UI for the shopping-list skill. Adds a /shopping page to Second Brain with full CRUD — view, add, edit, check off, and delete items. Requires the shopping-list skill to be installed (shares the same data files). --- # Shopping List UI Web interface for managi... | [] | clean | high | gpt-5.5 | This skill transparently adds a shopping-list web page that edits the expected shared shopping-list data. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:02:59.245000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
be3bcdac9ab1ec03a707ffc8f2f56783b44792dbd88e0a8e2a879669929eae23 | mikaelaldy/x-analytics-loop | 1.0.0 | --- name: x-analytics-loop description: Review X performance and choose next experiments. Use when analyzing weekly post metrics, diagnosing winners and losers, or planning what to repeat. --- # X Analytics Loop Run this weekly. ## Collect For each post, ask for or infer: - date/time - topic - format - hook - impressio... | [] | clean | high | gpt-5.5 | This is an instruction-only skill for reviewing X post metrics and planning content experiments, with no hidden automation or privileged access. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-16T03:43:16.346000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
1dac9f0c05ef7d3aefd1958b17e900177ec098acf0e4c347e67bef3f21f31f86 | alirezarezvani/email-template-builder | 1.0.0 | --- name: "email-template-builder" description: "Email Template Builder" --- # Email Template Builder **Tier:** POWERFUL **Category:** Engineering Team **Domain:** Transactional Email / Communications Infrastructure --- ## Overview Build complete transactional email systems: React Email templates, provider integration,... | [] | clean | high | gpt-5.5 | This is an instruction-only email template skill with purpose-aligned examples, though its sample tracking helper should be reviewed before production use. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 10 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:09:53.205000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
92bd13de11250d00d7ecb1c9e71e59cfef26b5073a19042c8de6539e1a3eab11 | maosangtianshi/zeh-hello-community | 0.1.1 | --- name: zeh-hello-community description: Generate short Chinese notices for community teachers, grid workers, and public service staff. --- # zeh-hello-community Use this skill when the user needs a short formal notice for community staff. ## Output format 1. 标题 2. 适用对象 3. 正文 4. 注意事项 ## Rules - Reply in Chinese. - Ke... | [
{
"path": "_meta.json",
"content": "{\n \"ownerId\": \"kn76fhpzxv6v6z61a15tq9jr8h82q99w\",\n \"slug\": \"zeh-hello-community\",\n \"version\": \"0.1.0\",\n \"publishedAt\": 1773286519275\n}\r\n",
"sha256": "179b4c112dff03e73383fee39bea6391f7471b66cda57cae0cc59075ca751d68",
"sizeBytes": 140
},
... | clean | high | gpt-5.5 | This skill is a narrow Chinese notice-writing helper with no hidden access, persistence, network use, or credential handling evident in its files. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:11:22.561000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
20bb77c21a7204ddd72f5ca9283a2fdd12d1568d47dcd7f72f3073df953590da | agistack/interview | 3.0.0 | --- name: interview description: Interview preparation system with company research, story building, and mock interview practice. Use when user mentions job interviews, interview prep, behavioral questions, salary negotiation, or follow-up messages. Researches companies, builds story libraries, runs mock interviews, pr... | [] | clean | high | gpt-5.5 | This interview-prep skill saves interview notes locally and does not show evidence of external sharing, credential access, destructive actions, or hidden execution. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 19 | LOW | 2 | [
"LP3",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:08:38.896000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
9f9eb91f4796ff47de19a3fd64227a73a85d01af80584912184f2f8bd4fd64a5 | zhengcookie/forklift-live | 1.0.0 | --- name: forklift-live description: "叉车配件直播运营助手，支持直播话术生成、询盘整理、客户跟进、风险预警。关键词：叉车配件、直播带货、叉车轮胎、叉车电瓶、叉车滤芯、叉车油封、叉车链条、叉车货叉、询盘转化、客户跟进、直播话术" --- # 叉车配件直播运营助手 ## 核心目标 1. 帮主播把直播间讲得更专业 2. 快速判断客户需求 3. 生成直播话术 4. 整理客户询盘 5. 提醒收集车型、吨位、年份、品牌、铭牌、旧件照片、尺寸等关键信息 6. 降低错配、误配、售后纠纷风险 7. 生成直播后的私信、微信、社群跟进文案 ## 重要原则 - 不乱报价格、不乱承诺现货、不乱承诺通用、不乱承诺质保、不乱... | [
{
"path": "references/inquiry-forms/customer-inquiry.md",
"content": "# 📋 叉车配件客户询盘信息收集表\n\n## 基础信息\n\n| 字段 | 内容 | 话术引导 |\n|------|------|----------|\n| 客户称呼 | ______ | \"老板怎么称呼？\" |\n| 联系电话 | ______ | \"留个电话，有货第一时间通知您\" |\n| 所在地区 | ______ | \"您是在哪个城市？方便算运费\" |\n| 公司/车队名称 | ______ | \"方便问一下您公司或车队名称吗？\" |\n\... | clean | high | gpt-5.5 | This is a markdown-only forklift-parts livestream sales assistant with ordinary customer-inquiry privacy considerations but no hidden execution, persistence, or unsafe behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 20 | LOW | 2 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-17T08:21:25.059000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
831019d8e7c4163292bc87e2a881bd2f70121c5ed7049b02a897389e9e65d6bd | expysf98/cloudflare-image-gen | 1.0.0 | --- name: cloudflare-image-gen description: Generate images using Cloudflare Workers AI flux-1-schnell model. Use when user requests image generation with text-to-image using Cloudflare Workers API. --- # Cloudflare Image Generation This skill generates images using Cloudflare Workers AI `@cf/black-forest-labs/flux-1-s... | [] | suspicious | high | gpt-5.5 | This image-generation skill is review-worthy because it exposes a Cloudflare credential and lets image prompts flow into a shell command. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 60 | suspicious | 100 | CRITICAL | 7 | [
"AST4",
"LP3",
"SDI-2",
"SQP-2",
"TP4"
] | [
"Dangerous Code Execution",
"MCP Least Privilege",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:02:01.056000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
406dd566257e3070a0069ba5ec6de85a194926bb52893e06689679006ad74d40 | davidharket/clawng-term-memory | 1.2.0 | --- name: clawng-term-memory description: Make your OpenClaw agent portable and persistent. Version-controls SOUL.md, MEMORY.md, and all core knowledge files using git with automatic GitHub push — so your agent's identity, memory, and operating rules are backed up and can be restored on any machine by cloning the repo.... | [] | suspicious | high | gpt-5.5 | This skill openly backs up agent memory to GitHub, but it gives itself broad automatic authority over sensitive agent files with limited review safeguards. | clean | 0 | [] | clean | 0 | 0 | 0 | 59 | suspicious | 100 | CRITICAL | 7 | [
"SDI-2",
"SQP-1",
"SQP-2",
"SSD-3"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:01:37.373000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
873a0c3ce3993eb840d7e2956ed550b15727924db205e8e42221c539e26a6216 | clawto/wechat-manager | 1.0.0 | --- name: wechat-manager description: "微信智能管家：自动回复、消息分类、聊天记录分析、联系人管理。Trigger on: 微信, WeChat, 群聊, 朋友圈, 公众号, 聊天记录, 联系人, 表情包, 小程序, 红包." version: 1.0.0 license: MIT --- # WeChat Manager 💬 微信智能管家 — 通过微信官方 API 实现消息管理、自动回复和数据洞察的 OpenClaw Skill。 ## What it does - **消息管理**: 收发消息、群聊管理、自动回复规则 - **联系人管理**: 搜索联系人、分组管理、标签系统 - **聊天分... | [] | suspicious | high | gpt-5.5 | This WeChat skill is purpose-built for chat management, but it needs Review because it handles private messages, contacts, monitoring, and auto-reply rules without clear consent or safety limits. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 100 | CRITICAL | 7 | [
"LP3",
"SQP-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-28T03:59:57.904000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
23dc6c020b3b0f036dcb764b29e894e96c8f29a9b765d93faabe4ae8abc79e3c | alirezarezvani/quality-manager-qmr | 2.1.1 | --- name: "quality-manager-qmr" description: Senior Quality Manager Responsible Person (QMR) for HealthTech and MedTech companies. Provides quality system governance, management review leadership, regulatory compliance oversight, and quality performance monitoring per ISO 13485 Clause 5.5.2. triggers: - management revi... | [] | clean | high | gpt-5.5 | This skill is a coherent ISO 13485 quality-management helper with an optional local reporting script and no evidence of hidden access, persistence, or data exfiltration. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:56:00.410000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
be30e1ab148e90bacedc7e6dc74cb04e210f17b1a60d3245d3e9cf7a7d1c4238 | membranedev/desktime | 1.0.1 | --- name: desktime description: | DeskTime integration. Manage Users, Organizations, Projects, Tasks, Reports, Absences. Use when the user wants to interact with DeskTime data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT metadata: author: membrane version: "1.... | [] | suspicious | high | gpt-5.5 | This DeskTime skill is a coherent integration, but it gives an agent broad authenticated access to sensitive workplace data without enough limits or confirmation guidance. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | suspicious | 30 | MEDIUM | 3 | [
"SDI-2",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T15:23:44.991000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
ab6e777c2f076e043fed6f78c7aaa3ff032d27cb65818d5f47d39056514fbb25 | rajniwebdeveloper/challanx | 1.0.3 | --- name: challanx version: 1.0.1 author: "ChallanX Team <[REDACTED_SECRET]>" description: "All-in-one RTO, challan, RC lookups and media downloader features for ChallanX. Accepts a single input — a URL, a free-text message, or an image — and returns concise, actionable results (download links, summaries, or document a... | [] | clean | high | gpt-5.5 | This is a coherent ChallanX API integration, but users should understand that submitted URLs, text, images, and API-key-authenticated requests go to ChallanX. | clean | 0 | [] | clean | 0 | 0 | 0 | 36 | suspicious | 71 | HIGH | 4 | [
"SQP-1",
"SQP-2",
"TP4"
] | [
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:40:13.981000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
9f5cf8e0dd4f173c837660ac2b61e9912f880f8a5a8dff5acff24ee5fa7cbf55 | jason513597/clawpilot | 1.0.0 | --- name: clawpilot description: > Expert skill for OpenClaw (v2026.2.19) — self-hosted AI gateway connecting chat apps (WhatsApp, Telegram, Discord, Slack, iMessage, Signal, LINE, Matrix, Teams, Google Chat, BlueBubbles) to AI agents. Use when user asks about: (1) Installing, configuring, or updating OpenClaw, (2) Set... | [] | suspicious | high | gpt-5.5 | ClawPilot is a coherent OpenClaw helper, but it needs review because it can run local/network checks and inspect sensitive OpenClaw configuration and session data. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 78 | HIGH | 6 | [
"LP3",
"SQP-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:08:46.456000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
07f0cc0fb2e7c04789b90f700cc138d45314603af319b6407d1838125ba58222 | yo-onething/laike-sales | 1.1.0 | --- name: laike-sales description: 来客推商城系统。用于回答商城源码相关咨询、产品介绍、功能解答。当用户咨询商城源码、电商系统、软件开发等问题时触发。 --- # 来客推商城系统 ## 公司信息 - **官网**：https://www.laiketui.com - **客服微信/QQ**：13237476750 - **联系电话**：13237476750 ## 产品矩阵 | 产品 | 说明 | 演示站 | |------|------|--------| | V3 | 多用户商城系统 | https://pc.thinkinshop.laiketui.net/ | | ThinkInShop |... | [] | clean | high | gpt-5.5 | This is a simple product-information skill for a Chinese e-commerce system, with no code execution, credential access, persistence, or hidden data handling. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 10 | LOW | 1 | [
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:34:38.603000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
6f6fa381d8385c1346eccc88c42a54d62184e470306d4c8b94d7070ad5056486 | 734818028/keep-record | 1.5.8 | --- name: keep-record description: > Assist users in recording personal health data, including diet, weight, body fat, body measurements, exercise, sleep, and menstrual cycle. Activate when the user clearly expresses an intent to record, log, check in, or save health-related information. Examples that should trigger in... | [] | suspicious | high | gpt-5.5 | This appears to be a real Keep health-recording skill, but it has broad automatic activation, persistent local account tokens, install-time side effects, and loosely scoped file upload behavior that users should review carefully. | clean | 0 | [] | clean | 0 | 0 | 0 | 39 | suspicious | 100 | CRITICAL | 19 | [
"LP3",
"SDI-1",
"SDI-2",
"SQP-1",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-11T08:45:31.360000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
93b21ff168aa94830b218ea2b704daa9df7288c9aa505010c02b23e54f603095 | neldar/webchat-voice-proxy | 0.2.3 | --- name: webchat-voice-proxy deprecated: true description: > ⚠️ DEPRECATED — This skill has been split into two separate skills for better modularity: **webchat-https-proxy** (HTTPS/WSS reverse proxy) and **webchat-voice-gui** (mic button, VU meter, i18n). Install those instead. Use **webchat-voice-full-stack** for on... | [] | clean | high | gpt-5.5 | The skill matches its stated local voice-input purpose and discloses its persistent local proxy, UI injection, and token-authenticated transcription behavior. | suspicious | 1 | [
"suspicious.dangerous_exec"
] | clean | 0 | 0 | 0 | 65 | suspicious | 100 | CRITICAL | 6 | [
"RA2",
"SQP-2",
"TT3"
] | [
"Rogue Agent",
"Data Flow"
] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.dangerous_exec"
],
"summary": "Detected: suspicious.dangerous_exec",
"checked_at": "2026-05-10T06:56:14.653000Z",
"scanner_version": "v2.4.23",
"findings": [
{
"code": "susp... | train |
ce4642680b54656fbb0e2042991f8217486f3cb2e8f0a941a652490f2a18bed4 | zarruk/wa-relay-skill | 0.2.1 | --- name: wa-relay description: "WhatsApp message relay for OpenClaw v0.2.0. Routes third-party DMs to a relay agent that forwards them to the main agent via sessions_send. The main agent notifies the owner with a suggested response. Supports direct number allowlists." --- # wa-relay v0.2.0 Routes third-party WhatsApp ... | [] | suspicious | high | gpt-5.5 | This WhatsApp relay is mostly transparent and purpose-aligned, but it deserves Review because setup copies main-agent credentials and patches core OpenClaw files. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 64 | suspicious | 100 | CRITICAL | 9 | [
"SDI-1",
"SDI-2",
"SDI-4",
"SQP-2",
"TP4"
] | [
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:59:17.596000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
562b1b0b47bdb582fabdd1370f169860a119b066273acffdfaf1eed1c3259b43 | marlowne12/pinterest-trends-dashboard | 1.0.0 | --- name: pinterest-trends-dashboard description: > Real-time visualization dashboard showing trending ClickBank niches, Pinterest search volume estimates, and affiliate payout potential. Helps users identify high-opportunity affiliate niches for Pinterest-based marketing. Part of Max's ClickBank + Pinterest automation... | [
{
"path": "BRIEF_PRD_ARCH.md",
"content": "# Pinterest Trends Dashboard — Brief + PRD + Architecture\n\n**Created:** 2026-03-08 22:14 PT \n**Status:** Ready for Review \n**Author:** Max (Autonomous Build)\n\n---\n\n## Executive Summary\n\nA real-time visualization tool showing trending ClickBank niches, r... | clean | medium | gpt-5.5 | The skill appears purpose-aligned for affiliate marketing analytics, with privacy and third-party script caveats but no evidence of hidden or destructive behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 20 | LOW | 2 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:26:05.553000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
cdbd9192a0305b6f62e3c2fb924a3a6da15258d96acf4d8208f8b75e1483858b | wangzhiming1999/felo-twitter-writer | 1.0.0 | --- name: felo-twitter-writer description: "Dual-mode Twitter/X writing tool. Mode 1: input a Twitter account, auto-fetch popular tweets and extract a writing style DNA document. Mode 2: based on style DNA and a topic, compose high-quality tweets, threads, or X long-form posts. Use when users want to analyze Twitter st... | [
{
"path": "clawhub.json",
"content": "{\r\n \"name\": \"Felo Twitter Writer\",\r\n \"tagline\": \"Analyze tweet style DNA and compose tweets, threads, X posts with brand style in Claude Code\",\r\n \"description\": \"Felo Twitter Writer lets you analyze any Twitter/X account's writing style and extract a... | suspicious | medium | gpt-5.5 | The skill’s Twitter writing workflow is coherent, but it automatically reuses and persists work into an existing private Felo LiveDoc without clear user confirmation or isolation. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | clean | 10 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:32:06.283000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
58fb3a331ec9c635f072502ce4127bee39cdaaaef46890bbb95ab501565567b5 | tetsuakira-vk/japanese-daily-drill | 1.0.0 | --- name: Japanese Daily Drill slug: japanese-daily-drill description: Generates a personalised Japanese language practice session based on JLPT level. Covers vocabulary, grammar, kanji, reading, and speaking prompts. Fresh content every session. version: 1.0.0 author: tetsuakira-vk license: MIT tags: [japanese, jlpt, ... | [
{
"path": "README.md",
"content": "# Japanese Daily Drill — OpenClaw Skill\n\n**A complete, personalised Japanese practice session every time you ask.**\n\nLevel-appropriate, fresh content every session. No repetition, no subscriptions, no app to open.\n\n---\n\n## What it does\n\nTell it your level. Get ba... | clean | high | gpt-5.5 | This is a simple Japanese study prompt skill with no executable code, credentials, network access, or hidden data handling. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 5 | LOW | 1 | [
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:23:44.293000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
9d82ff3633fd9ec4d724794b4fe39088630255cf871dfc93d1c8794d986a0bff | justoneapi/justoneapi-tiktok-get-user-post | 1.0.0 | --- name: TikTok User Published Posts API description: Call GET /api/tiktok/get-user-post/v1 for TikTok User Published Posts through JustOneAPI with secUid. author: JustOneAPI homepage: https://api.justoneapi.com metadata: {"openclaw":{"homepage":"https://api.justoneapi.com","primaryEnv":"JUST_ONE_API_TOKEN","requires"... | [] | clean | high | gpt-5.5 | This is a narrow JustOneAPI helper for fetching TikTok published-post data, with a real but disclosed credential-handling caveat. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | suspicious | 30 | MEDIUM | 3 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T09:32:10.232000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
b0652a231b02375e7720ec53e98c8d91b6db63128508ba797b6b297a1cf262d4 | 892153301/ai-video-workflow | 1.0.0 | --- name: ai-video-workflow description: | AI视频完整制作工作流技能。适用于： 1. 用户要求"制作AI视频"、"做AI视频"、"AI视频创作" 2. 用户提到"视频工作流"、"视频制作流程" 3. 用户需要AI视频制作的系统化指导 核心：七步流程 + 打磨检查贯穿每步 + 提示词模板 --- # AI视频制作工作流 > 版本：v5.0 精简版 > 基于：v4.2 完整版 ## 触发场景 当用户有以下需求时自动执行： - "帮我制作一个AI视频" - "怎么做AI视频" - "从剧本开始制作视频" - "AI视频工作流" ## 核心架构 | 角色 | 功能 | |------|------... | [] | clean | high | gpt-5.5 | This is a markdown-only AI video workflow guide with no hidden code, credential use, or destructive behavior found. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 40 | MEDIUM | 4 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:22:47.467000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
cdefab221a6fc5d06ebcd4fc065230e318abc05396818f7bcbbbf13fed310d8e | techievena/eonik-creative-audit | 1.0.10 | --- name: "eonik creative audit" slug: "eonik-creative-audit" version: "1.0.9" description: "Identifies creative fatigue, budget leaks, and scaling opportunities by running the eonik creative audit engine." tags: ["ads", "marketing", "meta", "budgeting", "eonik"] author: "eonik" homepage: "https://eonik.ai" metadata: o... | [] | clean | high | gpt-5.5 | This is a disclosed Meta ads audit tool that handles sensitive ad-account data, but its access and outputs are mostly scoped and user-gated. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 65 | HIGH | 5 | [
"LP3",
"SDI-2",
"SDI-4",
"SQP-1",
"SQP-3"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-06T06:15:21.154000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
d42d4821b788008996ab073916335ad6287d7fda21a0f3ae2204de8636f29031 | gora050/generated-photos | 1.0.2 | --- name: generated-photos description: | Generated Photos integration. Manage Persons. Use when the user wants to interact with Generated Photos data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT homepage: https://getmembrane.com repository: https://github.com... | [] | clean | high | gpt-5.5 | This is a disclosed Generated Photos integration that uses Membrane for account connection and API access, with no executable skill payload. | clean | 0 | [] | clean | 0 | 0 | 0 | 54 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:16:02.827000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
b142edd02fb1f22b813a50e728e32b2bfc68902bb697d6ed4a98d7199f92d036 | jkfaris94/hybrid-gateway | 1.0.2 | --- name: hybrid-gateway description: Set up and troubleshoot a hybrid OpenClaw architecture where the gateway runs on a cloud VPS and a local machine (Mac Mini, desktop, Raspberry Pi, etc.) acts as a node. Covers Tailscale networking, gateway bind modes, node pairing, LaunchAgent/systemd auto-start, exec routing, SSH ... | [
{
"path": "README.md",
"content": "# Hybrid Gateway — VPS + Local Node for OpenClaw\n\nRun your OpenClaw gateway on a cloud VPS and connect a local machine (Mac Mini, desktop, Raspberry Pi) as a node — getting the best of both worlds.\n\n## Why?\n\nMost OpenClaw users start on a VPS. It's cheap, always onli... | suspicious | high | gpt-5.5 | This is an instruction-only OpenClaw setup guide, but it recommends network exposure and remote command execution settings that deserve careful review before use. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 30 | MEDIUM | 3 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:31:15.801000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
4e888233ca697061341c75263d836d03fdd5cd77e8fdbd54919c7e35af1d2c30 | 403914291/wechat-publisher-easy | 2.0.3 | # 微信公众号发布技能 **技能名称：** wechat-publisher **版本：** V2.0.0（块布局 v3.0） **描述：** 自动发布 AI 新闻到微信公众号草稿箱（块布局固定版） **作者：** 小蛋蛋 **技术支持：** [REDACTED_SECRET] **公众号：** 心识孤独的猎手 **微信：** 心识孤独的猪手 --- ## 📋 功能特性 - ✅ 自动收集 32 条 AI 新闻（4 类×8 条） - ✅ 自动生成 HTML 格式内容（块布局 v3.0） - ✅ 自动发布到公众号草稿箱 - ✅ **固定模板**：块布局 v3.0（官方唯一模板） - ✅ 50 次免费试用 + 8.8 元永久买断 - ✅... | [] | suspicious | high | gpt-5.5 | The skill appears to do its stated WeChat draft-publishing job, but it asks for powerful account credentials and documents unsafe ways to store and display them. | clean | 0 | [] | clean | 0 | 0 | 0 | 67 | suspicious | 100 | CRITICAL | 11 | [
"SDI-4",
"SQP-2",
"SSD-3"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:32:46.375000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
8d7cbbc6e85d762ab4a45934a3bfe0f6952f06508c81f9bafd48212dd7d622ad | jiuping520/baiyin-voice-generate-skill | 1.0.3 | --- name: baiyin-voice-generate-skill description: 使用百音开放平台创建 AI 语音任务，支持文本转语音、音色克隆，并在同一 skill 内继续查询任务状态和结果链接。用于用户要生成语音、克隆音色、查询语音任务进度或下载结果时。 --- > **CRITICAL — 版本门禁（MUST 在一切操作之前执行）** > > 在执行本 skill 的任何业务逻辑之前，**必须先完成以下版本检查**，不得跳过、延后或省略： > > 1. 读取本地 `_meta.json`，取得 `ownerId`、`slug`、`version`、`publishedAt`。 > 2. 请求 SkillHu... | [
{
"path": "_meta.json",
"content": "{\n \"ownerId\": \"baiyin\",\n \"slug\": \"baiyin-voice-generate-skill\",\n \"version\": \"1.0\",\n \"publishedAt\": 1775791475000\n}\n",
"sha256": "45e8987a840c4c465b59c2b9beaf3d704a13013f7dcc08cbb19c022dd68facc1",
"sizeBytes": 119
}
] | suspicious | high | gpt-5.5 | The voice-generation workflow mostly matches its purpose, but the skill also requires a silent pre-task self-update check that can change the installed skill without clear user approval. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 45 | MEDIUM | 3 | [
"SDI-2",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:38:51.177000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
83fbd816c9263e17f03bfae64d7faf46a7c625bffc6bc4ee235cef8959d5b7a1 | enderfga/ssh-lab | 0.1.0 | # ssh-lab — Remote Server SSH Workbench > Structured SSH operations for managing remote GPU servers. > Standardizes what's universal (GPU/disk/process), delegates interpretation to the LLM. ## When to Use - User asks about GPU status, server health, remote processes - Need to check if training is running, disk is full,... | [] | suspicious | high | gpt-5.5 | This is a mostly coherent SSH server-management skill, but it gives agents broad remote command and file-transfer power with weak scoping and a local command-injection risk in host resolution. | suspicious | 3 | [
"suspicious.dangerous_exec"
] | clean | 0 | 0 | 0 | 64 | suspicious | 78 | HIGH | 6 | [
"SDI-2",
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.dangerous_exec"
],
"summary": "Detected: suspicious.dangerous_exec",
"checked_at": "2026-05-10T07:18:18.776000Z",
"scanner_version": "v2.4.23",
"findings": [
{
"code": "susp... | train |
69af2e5ec2a2638d1d21b0f83c72ac0f9204766b6a4bb396e89ad549f62be29d | membranedev/caspio | 1.0.2 | --- name: caspio description: | Caspio integration. Manage data, records, and automate workflows. Use when the user wants to interact with Caspio data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT homepage: https://getmembrane.com repository: https://github.com... | [] | suspicious | high | gpt-5.5 | This Caspio skill is a coherent integration, but it gives an agent broad authenticated ability to change business data without clear confirmation boundaries. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 10 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T15:16:58.809000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
3b8706d2b45109cdcbd9f9e05201fcc6188bddb85aa402f7bfa0300f83004981 | temrjan/rustok-wallet | 0.2.2 | --- name: rustok-wallet description: Self-custody Ethereum Agent Wallet. All supported chains enabled by default (Ethereum, Arbitrum, Base, Optimism, zkSync, Sepolia, Arbitrum Sepolia). User assumes all risks. Runs locally. Preview/execute ETH sends with hard policy limits, track DeFi positions. All actions are append-... | [
{
"path": "claw.json",
"content": "{\n \"name\": \"rustok-wallet\",\n \"version\": \"0.2.2\",\n \"description\": \"Self-custody Ethereum Agent Wallet — all supported chains enabled by default (Ethereum, Arbitrum, Base, Optimism, zkSync, Sepolia, Arbitrum Sepolia). User assumes all risks. Preview and exec... | suspicious | high | gpt-5.5 | This is a disclosed local Ethereum wallet skill, but it gives an agent real fund-moving authority with permissive defaults that users should review carefully. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | suspicious | 100 | CRITICAL | 7 | [
"E1",
"LP3",
"TM2"
] | [
"Data Exfiltration",
"MCP Least Privilege",
"Tool Misuse"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-27T19:42:56.389000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
b86be147834c52471bf34782d21f5d35b7f3635f87ec733b1fd75a9a33b7e83a | kalvinrv/nano-banana-edit | 0.1.1 | --- name: nano-banana-edit displayName: "🫧 Nano Banana Edit — Pro Pack on RunComfy" description: > Edit images with Google Nano Banana 2 (image-to-image edit endpoint) on RunComfy. Documents Nano Banana Edit's strengths (preserve subject identity, swap background, localize edits with spatial language, multi-image batc... | [] | clean | medium | gpt-5.5 | The artifacts are coherent developer/maintainer skills for ClawHub and Convex work, with sensitive actions mostly disclosed and gated by user intent. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-29T12:04:51.895000Z",
"scanner_version": "v2.4.2",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
e546c9ebe2a16ef788fd7514ce765fda6021391161807b76c40fb3e0e9177671 | francemichaell-15/video-compressor-bad-quality | 1.0.0 | --- name: video-compressor-bad-quality version: "1.0.0" displayName: "Video Compressor Bad Quality Fix — Restore Quality From Compressed Videos" description: > Skip the learning curve of professional editing software. Describe what you want — fix the blocky artifacts and restore sharpness after compression — and get re... | [] | clean | high | gpt-5.5 | This is a cloud video-restoration/editing skill that is broadly scoped but disclosed and does not show hidden installation, credential theft, destructive behavior, or unrelated local access. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 75 | HIGH | 6 | [
"SDI-1",
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:37:50.762000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
b0b5caded65f062dcc06f06e16fcf1c2c65593a15e68df72c3cbbeeda903b83b | zhoubo416/invt-automation | 1.0.0 | --- name: invt-automation version: 1.0.0 author: zhoubo416 description: > Use this skill when users ask about industrial automation, variable frequency drives (VFDs), servo systems, controllers (PLC/HMI), motion control, woodworking equipment, CNC routers, edge banders, panel saws, engraving machines, production lines,... | [] | clean | high | gpt-5.5 | This is a markdown-only INVT sales referral skill with promotional bias but no code execution, data access, or persistence. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 20 | LOW | 2 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:45:43.994000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
1540fba4df9aafc07f0d3b6e9d18afae5752c6b071d41a0d7738e5a990e0fd8d | meirk-brd/clearweb | 1.0.0 | --- name: clearweb description: | Complete web access for AI agents via Bright Data CLI. Replaces native web_fetch, web_search, and browser tools with reliable, unblocked access to the entire web. USE FOR: Any web task — searching Google/Bing/Yandex, reading any webpage, extracting structured data from Amazon/LinkedIn/... | [
{
"path": "references/web-search.md",
"content": "# Web Search Reference\n\nComplete reference for web search operations via `bdata search`.\n\n## Command Syntax\n\n```bash\nbdata search \"<query>\" [options]\n```\n\n## All Options\n\n| Flag | Description | Default |\n|------|-------------|---------|\n| `--... | suspicious | high | gpt-5.5 | ClearWeb is a disclosed Bright Data web-scraping skill, but it gives agents very broad anti-bot scraping and credentialed external-service access with weak usage boundaries. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 50 | MEDIUM | 5 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:22:50.181000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
c007523414370dccf8c0a58e5ae3bb7c0aaf0d39e9e34d7614d322523b115d8e | dagangtj/multi-bounty-scanner | 1.0.0 | # Multi-Platform Bounty Scanner ## Overview Automatically scan 50+ bug bounty and OSS bounty platforms for new opportunities. Save hours of manual searching. ## When to use Use this skill when: - You want to find new bounty opportunities across multiple platforms - You need to automate daily bounty scanning - You want ... | [] | clean | high | gpt-5.5 | This is a local GitHub bounty-search helper with some overstated feature claims, but no hidden, destructive, or exfiltrating behavior was found. | suspicious | 1 | [
"suspicious.dangerous_exec"
] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.dangerous_exec"
],
"summary": "Detected: suspicious.dangerous_exec",
"checked_at": "2026-05-10T07:17:48.473000Z",
"scanner_version": "v2.4.23",
"findings": [
{
"code": "susp... | train |
630259ac2205d937c13fd250a73bd189b5e5b2e5fe31e29030fc19d060212c7c | zhaobod1/huo15-openclaw-memory-curator | 1.1.1 | --- name: huo15-openclaw-memory-curator version: 1.1.0 description: "记忆整理技能 — 审查结构化记忆，提取洞察，更新 MEMORY.md，清理过期条目。" homepage: https://github.com/zhaobod1/huo15-openclaw-enhance metadata: { "openclaw": { "emoji": "🧠", "requires": { "bins": [] } } } --- # 记忆整理 (Memory Curator) 定期整理和优化结构化记忆系统。 ## 使用时机 ✅ **使用此技能当：** - "整理一下记... | [] | clean | high | gpt-5.5 | This is a simple instruction-only memory cleanup skill that does what it says, though users should run it interactively because it can delete memories and update MEMORY.md. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 25 | MEDIUM | 3 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:37:27.204000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
8a9802f43f63dc39e56f92dcf584f62664d7ca57329df2e8db24cb71e5cd2559 | jamesjohnfox/tradekix | 1.0.0 | --- name: tradekix description: Query financial market data via the Tradekix API — stock prices, crypto, forex, indices, market news, earnings, economic events, Congressional trades, and social sentiment. Use when the user asks about markets, stock prices, trading data, economic calendars, or financial news. Also handl... | [] | suspicious | high | gpt-5.5 | Tradekix appears to be a legitimate market-data skill, but it needs review because it can automatically create an account, send user identity details, store an API key locally, and run account-changing actions without clear confirmation. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 66 | suspicious | 97 | CRITICAL | 8 | [
"LP3",
"SDI-1",
"SQP-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:58:27.233000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
8433c39aa7397392c05e4a96f1438d3a8ae5b253a41dbe4770034449e0489ddd | gora050/simpletexting | 1.0.4 | --- name: simpletexting description: | SimpleTexting integration. Manage Contacts, Campaigns, Keywords. Use when the user wants to interact with SimpleTexting data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT homepage: https://getmembrane.com repository: https... | [] | suspicious | high | gpt-5.5 | The skill is a coherent SimpleTexting helper, but it gives an agent broad power to send messages and alter contacts or campaigns without clear confirmation safeguards. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | clean | 10 | LOW | 1 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T17:42:28.634000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
3aa9994ef8e7720b2509937ef5820a2bbe43c4dbdc822312397672ba6efb3bb0 | whille/minimax-tts-cn | 1.0.0 | --- name: MiniMax TTS description: 调用 MiniMax 语音合成 API 生成语音。支持系统音色、克隆音色、流式/非流式输出。使用场景：用户需要生成高质量中文语音、语音合成、文本转语音。 homepage: https://platform.minimax.io/docs/api-reference/speech-t2a-http metadata: openclaw: emoji: 🎙️ requires: bins: [python3] env: [MINIMAX_API_KEY] pip: [requests] primaryEnv: MINIMAX_API_KEY envHelp: MI... | [] | clean | high | gpt-5.5 | This is a normal MiniMax text-to-speech skill that sends user-provided text to MiniMax and saves returned audio locally. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 13 | LOW | 1 | [
"E1"
] | [
"Data Exfiltration"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:04:01.837000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
ae7ac9643e312452d2b8c9a03034effb6b14927d94230abbc2a1161583bebc4b | mjquinlan2000/zoom-users-mcp | 1.0.0 | --- name: zoom-users-mcp description: MCP server for the Zoom Users API. Exposes Zoom Users' REST API as read-only MCP tools. metadata: { "openclaw": { "requires": { "bins": ["mcporter", "zoom-users-mcp"], "env": ["NODE_MCP_SECRET_KEY"] } } } --- # Zoom Users MCP Server Read-only MCP server for the Zoom Users API. ## U... | [] | clean | high | gpt-5.5 | This skill is a small, disclosed Zoom Users API MCP wrapper with no bundled executable code or hidden behavior in the submitted artifact. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:10:25.160000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
fcec6a99fdcfd9feb941cd2753dea0c22427f006e8da85097df3549f1538fe81 | jswortz/secret-manager | 1.0.0 | --- name: secret-manager description: Manage API keys securely via GNOME Keyring and inject them into OpenClaw config. homepage: https://github.com/openclaw/skills metadata: {"clawdbot":{"emoji":"🔐","requires":{"bins":["secret-tool","systemctl","python3"]},"install":[{"id":"bash","kind":"bash","bin":"secret-manager.sh... | [
{
"path": "secret-manager.sh",
"content": "#!/bin/bash\n# secret-manager\n# Securely manage API keys in GNOME Keyring and inject them into OpenClaw configuration.\n#\n# Usage:\n# secret-manager <KEY_NAME> [value]\n# secret-manager list\n# echo \"val\" | secret-manager <KEY_NAME>\n#\n# Configuration (E... | suspicious | medium | gpt-5.5 | This skill appears intended to manage OpenClaw API keys, but it handles secrets and controls services in ways users should review carefully before installing. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | suspicious | 100 | CRITICAL | 8 | [
"LP3",
"PE3",
"SDI-1",
"SDI-2",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"Privilege Escalation",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:57:14.209000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
44a336458dca5d0680b36deb2208afe9be8e07e1f1510517c93e254f292a8f1a | hgvgfgvh/multi-capability-bug-closure-en | 1.0.0 | --- name: multi-capability-bug-closure description: >- Unified bug investigation and closure by combining source code, database, server logs, and software platform query capabilities. Use when users require evidence-based conclusions from real data rather than static code analysis only. --- # Multi-Capability Bug Closu... | [] | clean | medium | gpt-5.5 | The skill content is coherent and purpose-aligned, with disclosed use of local development, review, Convex, GitHub, and ClawHub moderation tooling. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:21:55.073000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
8041e927cb4293bef71dd597c87b6693c34aff1bb0d2ba0ee866a33e1a0130f1 | codenova58/styling | 0.1.0 | --- name: "Styling" description: "提供专业穿搭、发型建议、化妆技巧及个人形象顾问筛选。 " --- # 变美造型 ## 筛选项 - 地理距离（步行/骑行/驾车范围） - 社交热度（人气榜/点评分/新开业） - 消费档次（实惠/中端/轻奢/体验型） ## 返回字段 - 实时排队情况/预约通道/电子票号 - 营业时间变动/停车信息/导航精准路径 - 社区真实图库/探店笔记/避坑清单汇总 ## 示例请求 - “如何快速使用 变美造型 处理 [具体场景任务] ？” - “变美造型 的 [核心功能] 有哪些最新玩法/优惠信息？” ## 更新频率 - 结合地理信息变动 with 点评社区新热度进行灵活调整 ##... | [] | clean | high | gpt-5.5 | This is an instruction-only beauty and styling guidance skill with disclosed local-service recommendation fields and no code, credentials, persistence, or automatic actions. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 30 | MEDIUM | 3 | [
"SDI-1",
"SDI-4",
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:17:50.346000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
9b7eccc196f7e144428ab8b08486ccfeac88dd12a12242bbfc64c972dc074435 | mory128/editor-combine | 1.0.0 | --- name: editor-combine version: "1.0.0" displayName: "Editor Combine — Merge Clips Into One Video" description: > Skip the learning curve of professional editing software. Describe what you want — combine these clips into one seamless video with smooth transitions — and get merged video file back in 30-60 seconds. Up... | [] | clean | medium | gpt-5.5 | This is a cloud video-editing skill that sends user-selected media to NemoVideo, with no local executable code or evidence of hidden or destructive behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 65 | suspicious | 40 | MEDIUM | 4 | [
"SDI-1",
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-11T15:20:55.329000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
63a1b578f89846d3ae65fe6c9f05e6b3c352951bd71b7799f947e4c032650b00 | kr1json/ai-quota-check | 1.0.1 | --- name: ai-quota-check description: "**DEFAULT quota checker** - Use this skill FIRST when user says '쿼타', '쿼터', 'quota', '쿼타확인', '쿼터확인', or asks about quotas. Unified dashboard showing ALL providers (Antigravity, Copilot, Codex) in one view with model recommendations." metadata: {"clawdbot":{"emoji":"🧮","requires":... | [
{
"path": "index.js",
"content": "#!/usr/bin/env node\r\n/**\r\n * Smart Router - Unified Quota Monitor & Model Recommender\r\n * \r\n * Features:\r\n * 1. Check login status for all providers\r\n * 2. Unified quota dashboard (Antigravity + Copilot + OpenAI Codex)\r\n * 3. Task-based model recommendation wi... | suspicious | high | gpt-5.5 | This quota checker is mostly purpose-aligned, but it reads local auth/session data and silently runs Codex to refresh quota information without clear opt-in. | suspicious | 1 | [
"suspicious.dangerous_exec"
] | suspicious | 0 | 0 | 0 | 66 | suspicious | 100 | CRITICAL | 8 | [
"LP3",
"P6",
"SDI-2",
"SQP-1",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"System Prompt Leakage",
"MCP Tool Poisoning"
] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.dangerous_exec"
],
"summary": "Detected: suspicious.dangerous_exec",
"checked_at": "2026-05-10T06:52:17.630000Z",
"scanner_version": "v2.4.23",
"findings": [
{
"code": "susp... | train |
418ca084345c5931d851f50cbc1b92dd190215319bc616d46e8f2ebf1396695a | randzero/step-asr | 1.0.0 | --- name: step-asr description: Transcribe audio files to text via Step ASR streaming API (HTTP SSE). Supports Chinese and English, multiple audio formats (PCM, WAV, MP3, OGG/OPUS), real-time streaming output, and terminology correction prompts. version: 1.0.0 metadata: openclaw: emoji: "\U0001F399" requires: bins: - p... | [] | clean | high | gpt-5.5 | This skill is a straightforward cloud transcription helper that uploads the user-selected audio file to StepFun’s ASR API and optionally saves the transcript locally. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | suspicious | 39 | MEDIUM | 3 | [
"LP3",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:04:51.489000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
5db645e4906d7492fb7dc0d63205e2b9fc4be333130399cfce008e4cc97af022 | millerallen98/flexible-web-tester | 1.0.2 | --- name: flexible-web-tester description: "智能 Web UI 测试工作台，支持 MCP 直接驱动和 Python 脚本驱动双模式，三种测试模式，强制人工确认，安全可控。" --- # Flexible Web Tester - 智能 Web UI 测试工作台 ## 阶段 0：MCP 环境自检 (Pre-flight Check) [强制逻辑] 当用户启动此技能时，你必须**立即**执行以下隐式环境检查： 1. **检查文件读写能力**：尝试列出或读取一个已知存在的本地文件，验证 File System MCP 是否可用 2. **检查终端执行能力**：执行 `echo "test"` 或... | [
{
"path": "README.md",
"content": "# Flexible Web Tester\n\n智能 Web UI 测试工作台，支持 MCP 直接驱动和 Python 脚本驱动双模式。\n\n\n\n![Platform](https://img.shields.io/badge/platform-OpenClaw-lightgrey.sv... | suspicious | high | gpt-5.5 | This web-testing skill is mostly coherent, but it asks for broad browser, terminal, filesystem, and login access without enough safeguards for credentials, saved artifacts, or real website actions. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | suspicious | 95 | CRITICAL | 7 | [
"SQP-1",
"SQP-2",
"SSD-3"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:17:18.729000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
8890ecda450087bd845c306c1dc0d93d4142e1620f09a10131f299e22376fdfb | feixuelingcloud/yuque-skill | 1.1.0 | --- name: yuque description: 通过语雀开放API对语雀(Yuque)文档与知识库执行端到端自动化操作 —— 创建/读取/更新/删除文档、管理知识库目录(TOC)、调整Markdown排版(字号、颜色、标题、表格、代码块、提示框)、全文搜索及批量导入导出。任何用户提到"语雀""yuque""知识库""文档目录""TOC""语雀文档格式调整""语雀批量导出"或上传语雀导出的Markdown文件,即使没有显式说"使用skill",也应主动调用此skill。 license: Complete terms in LICENSE.txt --- # 语雀(Yuque)自动化Skill 把语雀的开放API包装成一组开... | [] | suspicious | high | gpt-5.5 | This Yuque automation skill matches its stated purpose, but it can make live authenticated changes and deletions with broad activation and weak confirmation safeguards. | suspicious | 1 | [
"suspicious.destructive_delete_command"
] | clean | 0 | 0 | 0 | 66 | suspicious | 100 | CRITICAL | 10 | [
"LP3",
"SDI-4",
"SQP-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.destructive_delete_command"
],
"summary": "Detected: suspicious.destructive_delete_command",
"checked_at": "2026-05-10T06:35:44.397000Z",
"scanner_version": "v2.4.23",
"findings": [
... | train |
7017e1693e0ceb6a2c2ee2e1e29ea1acecf73d1bf5d493aebf7c7aa2a88d5692 | liulike/browser-local-chrome | 1.1.0 | # 浏览器自动化配置技能 配置 OpenClaw 使用本地 Chrome 调试模式进行浏览器自动化。 ## 功能 - ✅ 一键启动 Chrome 调试模式 - ✅ 自动配置 OpenClaw 浏览器设置 - ✅ 启用 SSRF 白名单（允许访问任意网站） - ✅ 检查浏览器状态 ## 使用方法 ### 一键配置（推荐） ```bash node scripts/setup.js ``` 这会自动完成所有步骤：启动 Chrome → 配置 OpenClaw → 验证状态 ### 分步配置 ### 1. 启动浏览器 ```bash node scripts/start-chrome.js ``` ### 2. 配置 OpenClaw `... | [
{
"path": "HYBRID-MODE.md",
"content": "# 浏览器混合方案 - 使用说明\n\n## 🎯 配置策略\n\n**默认：** 官方 Browser (OpenClaw-managed) \n**备用：** 本地 Chrome 技能 (browser-local-chrome)\n\n---\n\n## 📋 方案对比\n\n| 特性 | 官方 Browser | 本地 Chrome 技能 |\n|------|-------------|-----------------|\n| **配置文件** | `openclaw` | `local-chrome` |\n| *... | suspicious | medium | gpt-5.5 | The skill appears to be for browser automation, but it explicitly promotes bypassing SSRF/private-network protections and enabling powerful local Chrome control, so it needs careful review before installation. | suspicious | 6 | [
"suspicious.dangerous_exec"
] | clean | 0 | 0 | 0 | 65 | suspicious | 100 | CRITICAL | 8 | [
"SDI-2",
"SQP-2"
] | [] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.dangerous_exec"
],
"summary": "Detected: suspicious.dangerous_exec",
"checked_at": "2026-05-10T06:33:23.910000Z",
"scanner_version": "v2.4.23",
"findings": [
{
"code": "susp... | train |
b4fe2b686aa9dec5ba9ac82b063ae1334f6b98e29961a047a666179eaaa66a9b | afrexai-cto/afrexai-client-onboarding | 1.0.0 | # Client Onboarding Checklist Generator Generate a comprehensive, customized client onboarding checklist for any service business. Covers intake, setup, communication, milestones, and success metrics. ## When to Use - Starting a new client engagement - Standardizing onboarding across your team - Reducing client churn f... | [] | clean | high | gpt-5.5 | This skill is a simple checklist-writing prompt with a visible promotional link and no code execution, local access, persistence, or hidden behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:14:26.765000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
6e8741fcf1606a398ccbf006b11e7a0e3cf3a938dc73938383da7532442d1efa | benson126/design-inspiration-collector | 1.0.3 | --- name: design-inspiration-collector description: 双平台设计灵感收集技能。当用户需要设计参考、UI灵感、视觉创意时触发。用户提出设计方向（如"医疗App"、"移动端UI"、"金融Dashboard"等），技能负责：(1) 使用Tavily搜索Dribbble、Pinterest两个平台的官方搜索结果（不包含个人作品集、画板）(2) 整理内容并附上链接 (3) 输出 Markdown 文档，命名为"关键词+日期时间"格式 (4) 推荐其他相关方向（不带链接）。触发词：找灵感、收集灵感、设计参考、UI参考、视觉灵感、设计趋势、Dribbble、Pinterest。 --- # 双平台... | [] | clean | high | gpt-5.5 | This skill appears to do what it claims: collect design-inspiration links from Dribbble/Pinterest and save local report files, with privacy and file-writing caveats. | clean | 0 | [] | null | null | null | null | null | suspicious | 100 | CRITICAL | 7 | [
"LP3",
"SQP-1",
"SQP-2",
"TP4"
] | [
"MCP Least Privilege",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-21T07:48:16.230000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"skillspector": {
"status": "suspicious",
"verdict": "DO_NOT_INST... | train |
bf037cffafeb5a6f37e0d9a089a750ba200472233e8adbe969e24929cd799c73 | bytesagain1/tesla-commander | 3.0.1 | --- name: "tesla-commander" version: "3.0.1" description: "Reference tool for life — covers intro, quickstart, patterns and more. Quick lookup for Tesla Commander concepts, best practices, and implementation patterns." author: "BytesAgain" homepage: "https://bytesagain.com" source: "https://github.com/bytesagain/ai-ski... | [
{
"path": "scripts/script.sh",
"content": "#!/usr/bin/env bash\n# tesla-commander — Tesla Commander reference tool. Use when working with tesla commander in devtools contexts.\n# Powered by BytesAgain | bytesagain.com | [REDACTED_SECRET]\nset -euo pipefail\n\nVERSION=\"3.0.0\"\n\nshow_help() {\n cat << '... | suspicious | low | gpt-5.5 | The automated scanners show no issue, but the skill artifact and install metadata were not available in the workspace for a real coherence review. | clean | 0 | [] | clean | 0 | 0 | 0 | 58 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:10:56.162000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
df5686235446d930d51ee33a47980f9e9d74df8821ab69283bcae3124b48a7ed | gora050/manychat-integration | 1.0.4 | --- name: manychat description: | ManyChat integration. Manage Organizations. Use when the user wants to interact with ManyChat data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT homepage: https://getmembrane.com repository: https://github.com/membranedev/appli... | [] | suspicious | high | gpt-5.5 | This appears to be a legitimate ManyChat/Membrane integration, but it grants broad authenticated access to business messaging data, including raw write and delete API requests, without clear guardrails. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | clean | 20 | LOW | 2 | [
"SDI-1",
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T17:11:04.145000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
6a7aaa7f220a5d2213a5326260c05857529fbfce35f2e5e8a85007a575028ded | brianleach/metra | 1.0.0 | --- name: metra description: Chicago Metra commuter rail — real-time train arrivals, vehicle tracking, service alerts, and schedule info for all 11 Metra lines serving the Chicago metropolitan area. Use when the user asks about Metra trains, Chicago commuter rail, or specific Metra lines and stations. homepage: "https:... | [] | clean | high | gpt-5.5 | This skill is a disclosed Metra transit lookup tool that uses a Metra API key and local GTFS cache in ways that match its purpose. | suspicious | 1 | [
"suspicious.env_credential_access"
] | clean | 0 | 0 | 0 | 56 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.env_credential_access"
],
"summary": "Detected: suspicious.env_credential_access",
"checked_at": "2026-05-10T06:57:43.035000Z",
"scanner_version": "v2.4.23",
"findings": [
{
... | train |
7e5614567895e5688d88cbb57148c9232fd03f6e026e3f907619027c0b03b4b4 | jack123255829/uncertainty-verification | 1.0.0 | --- name: fact-check description: | 事实核查与幻觉检测。在生成内容后检测并纠正可能的幻觉、编造或不准确信息。 触发条件：(1) 生成技术内容后 (2) 提供统计数据时 (3) 引用人物/事件时 (4) 生成代码示例后 --- # Fact-Check ## 核心原则 **验证结果 > 报告完成** 22%的"完成"任务实际失败，需要独立验证。 ## 触发条件 以下情况应触发事实核查： 1. 生成技术内容后 2. 提供统计数据或数字 3. 引用人物、事件、日期 4. 生成代码示例后 5. 完成外部操作后 ## 验证协议 ### 1. 结果验证 vs 动作验证 | 类型 | 说明 | 例子 | |--... | [] | clean | high | gpt-5.5 | This is a markdown-only fact-checking skill that asks the agent to verify claims and generated work, with no hidden code or installation behavior found. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:09:40.444000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
2411e4b8083b0d574be06f9445f9e919667353ea4c39d2306e2114d08b2607d3 | 1688aiinfra/report-to-visualization | 0.1.0 | --- name: report-to-visualization description: 将文本报告转换为可视化组件流。当用户需要将报告数据可视化或提到"可视化报告"、"用图表展示"、"图文并茂"时使用。 --- # 报告转可视化组件流 ## 任务目标 将文本报告数据转换为可视化组件配置，输出严格的MODULE格式。 ## 输入格式 用户需提供以下变量： - `${moduleData}`: 报告数据内容 ## 图表库规范 ### 1. 折线图 (Chart.Line) **使用条件**: 展示数据趋势，需至少5个连续、真实的时间点原始观测值 **数据格式**: ```json { "type": "Chart.Line", "... | [] | clean | high | gpt-5.5 | This is a markdown-only report-formatting skill that converts provided report text into visualization JSON and does not show hidden system access or unsafe behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 20 | LOW | 2 | [
"SQP-1",
"SQP-3"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-09T10:10:27.915000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
96a4015a75c55cd1edb1cc020931f3bb93c76a0632e66d782a409aa9658fa1c4 | tincomking/project-mode | 1.0.1 | --- name: project-mode description: "激活项目制多智能体开发协议。用于处理复杂的代码开发、系统搭建等需求。该工具将在后台自动拆解任务、调度程序员和测试员、更新 dev_project.md 并处理错误重试。" metadata: { "openclaw": { "emoji": "📋", "requires": {} }, } --- # 项目制多智能体开发协议 (Project Mode Skill) ## 概述 本 Skill 执行项目制多智能体开发协议，用于处理复杂的代码开发、系统搭建等需求。 ## 触发条件 当用户说"项目制"去执行某项任务时，必须激活本协议。 ## 执行流程 ### 阶... | [
{
"path": "project_mode_skill.py",
"content": "#!/usr/bin/env python3\n\"\"\"\n项目制多智能体开发协议 - 核心执行脚本\nProject Mode Skill Core Execution Logic\n\n用于处理复杂的代码开发、系统搭建等需求。\n自动拆解任务、调度程序员和测试员、更新 dev_project.md 并处理错误重试。\n\"\"\"\n\nimport json\nimport os\nimport re\nimport time\nfrom datetime import datetime\n\n# ====... | clean | high | gpt-5.5 | This is a disclosed project-workflow helper that persists local project tracking state, with no evidence of exfiltration, deception, or destructive behavior. | clean | 0 | [] | suspicious | 0 | 0 | 0 | 61 | suspicious | 39 | MEDIUM | 3 | [
"LP3",
"SQP-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:02:03.976000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "suspicious",
"verdict": "suspicious",
... | train |
f44fb611762d320a3e58162bf9f368e97c7d80166a96456935811fa6b64a8514 | subaru0573/agentic-workflow-automation-p | 1.0.0 | --- name: agentic-workflow-automation-p description: Design and orchestrate robust multi-step agent workflows with reusable blueprints. Automate complex trigger-action sequences, define deterministic workflows, and generate production-ready handoff artifacts. Ideal for streamlining automation pipelines, reducing manual... | [] | clean | high | gpt-5.5 | This is a local workflow-blueprint generator with one misleading dry-run flag, but no evidence of hidden network access, credential use, or unsafe persistence. | clean | 0 | [] | clean | 0 | 0 | 0 | 59 | suspicious | 26 | MEDIUM | 2 | [
"LP3",
"SDI-4"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:45:27.700000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
1fb11c7968011ca312a580685842641e7cf8fa9cf4c2ed917f9c934a6722c704 | swader/agent-skills-audit | 0.1.0 | --- name: audit-code description: Run a two-pass, multidisciplinary code audit led by a tie-breaker lead, combining security, performance, UX, DX, and edge-case analysis into one prioritized report with concrete fixes. Use when the user asks to audit code, perform a deep review, stress-test a codebase, or produce a ris... | [] | suspicious | medium | gpt-5.5 | The code-audit skill is coherent, but its optional sync script can overwrite or delete local agent skill directories and does not safely constrain destination names. | clean | 0 | [] | clean | 0 | 0 | 0 | 51 | suspicious | 52 | HIGH | 4 | [
"SDI-2",
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T04:57:38.043000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"verdict": "benign",
"reaso... | train |
056056f4861cac0e1b45adf8d437551f2e497f5c939f0d98c4e11a89b98bd168 | athola/nm-abstract-subagent-testing | 1.8.5 | --- name: subagent-testing description: Test skills via RED/GREEN/REFACTOR TDD with fresh subagents version: 1.9.5 triggers: - testing - validation - TDD - subagents - fresh-instances - validating skill behavior - preventing priming bias metadata: {"openclaw": {"homepage": "https://github.com/athola/claude-night-market... | [] | clean | medium | gpt-5.5 | This appears to be a testing-focused skill with a broad trigger-wording issue, but no evidence of hidden, destructive, or sensitive behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 10 | LOW | 1 | [
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-09T02:15:16.158000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
b4c140ff27232f9e9102b030a3ecb285e6703b89e027f596fb10367a0f7b9c84 | gora050/hankoio | 1.0.4 | --- name: hankoio description: | Hanko.io integration. Manage data, records, and automate workflows. Use when the user wants to interact with Hanko.io data. compatibility: Requires network access and a valid Membrane account (Free tier supported). license: MIT homepage: https://getmembrane.com repository: https://githu... | [] | suspicious | high | gpt-5.5 | This is a coherent Hanko.io integration, but it gives broad authenticated API power through Membrane, including write and delete requests, without clear built-in approval rules. | clean | 0 | [] | clean | 0 | 0 | 0 | 63 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T16:55:09.378000Z",
"scanner_version": "v2.4.22",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"check... | train |
979164f314eaebefff037ed95d73649d71490420524352599015c560fa2acbb3 | vladislavshad/workflow-automation-evm-wallets | 1.0.1 | --- name: ditto-workflow description: Creates, configures, and deploys on-chain automation workflows using the Ditto Network SDK. Use when the user asks to "create a workflow", "automate on-chain", "schedule transactions", "deploy a workflow", "set up recurring transfers", "swap tokens on a schedule", "automate DeFi", ... | [] | suspicious | high | gpt-5.5 | This looks like a legitimate Ditto workflow helper, but it handles raw wallet private keys and production crypto automation with too little safety framing. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 20 | LOW | 2 | [
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:13:36.805000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
8017ea5dbff482f50bec9aa4a4123a48074f17dc86e4802a3d6458489896aa33 | sarkcesscrewpay/ipo-tracker | 1.0.0 | --- name: ipo-tracker description: Tracks recent IPOs and generates TradingView watchlists. Auto-triggers on phrases like "recent IPOs", "tradingview watchlist", "IPO calendar", "IPO tracker", "IPO report", "new stock listings", or "IPO data". allowed-tools: - Bash - Read - Write metadata: mcpmarket-version: 1.0.0 --- ... | [] | clean | medium | gpt-5.5 | This appears to be a coherent IPO watchlist/reporting skill with routine command, dependency, file-output, and market-data retrieval behavior for its stated purpose. | clean | 0 | [] | null | null | null | null | null | clean | 15 | LOW | 2 | [
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-19T18:36:45.742000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"skillspector": {
"status": "clean",
"verdict": "SAFE",
"chec... | train |
83b35fe8ee7cb42f826cd28a033d6a82cd0c766518fc38192b10b765cae74592 | jcools1977/context-resume | 1.0.0 | --- name: context-resume version: 1.0.0 description: > Reconstructs your full working context when you return to a task after an interruption — meeting, lunch, sleep, weekend, or three-week vacation. Not session restore. Mental state restore. Answers the question every developer asks ten times a week: "Where was I?" au... | [] | clean | high | gpt-5.5 | This instruction-only productivity skill reviews local development context to help resume work, with disclosed but privacy-sensitive inputs and no evidence of hidden execution or data exfiltration. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | clean | 20 | LOW | 2 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:06:13.636000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
4041c8a25bc1ed02f63f692bfa69e8ef4e9a02da6e45947fd8dc1b55e641ff7d | goog/udon | 1.0.0 | --- name: udon description: > Search and browse 100 curated Japanese udon noodle recipes from Cookpad. Use when the user asks about udon recipes, wants udon dish ideas, searches for udon by ingredient (e.g. curry, miso, egg, kimchi, salmon), or mentions udon cooking. Covers grilled udon (yaki-udon), cold udon, nabeyaki... | [] | clean | high | gpt-5.5 | This skill is a straightforward local udon recipe search helper with no evidence of hidden access, persistence, credential use, or unsafe behavior. | clean | 0 | [] | clean | 0 | 0 | 0 | 64 | clean | 0 | LOW | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:32:16.050000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
fa6f35d9a454533078109bf150853b06ad90f498b3d14aac87090263fd214777 | 1688aiinfra/ozon-product-selection | 1.0.1 | --- name: ozon-product-selection category: official-1688 description: >- Ozon电商选品工具：根据用户指定的品类，自动搜索市场趋势、提取热门细分关键词、 调用1688搜索API逐一选品，按筛选条件过滤后输出每个细分市场的最优商品推荐。 触发词：寻找Ozon蓝海商品、Ozon潜力爆品、Ozon可铺货商品、Ozon可卖的货盘、 Ozon选品、帮我选品、找Ozon能卖的商品、Ozon热门商品推荐。 metadata: version: 1.0.1 label: 1688Ozon选品 author: 1688官方技术团队 --- # Ozon 电商选品 根据用户指定的... | [
{
"path": "requirements.txt",
"content": "requests>=2.20.0\nPyJWT>=2.0.0\n",
"sha256": "60d627e2b168242b7dfc8959ea0f9e17a7f30ef3875727860b05474756239681",
"sizeBytes": 30
}
] | clean | high | gpt-5.5 | This skill is a disclosed Ozon product-sourcing helper that uses AlphaShop/1688 APIs and API keys, with no evidence of hidden persistence, destructive behavior, or data exfiltration. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | suspicious | 100 | CRITICAL | 6 | [
"LP3",
"SC1",
"SC4",
"TP4"
] | [
"MCP Least Privilege",
"Supply Chain",
"MCP Tool Poisoning"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:22:03.218000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
a384bd59e9c10c061d69b48d184f406ebee1e94454f311507b350ac4c92eb7af | guitu917/ai-screen-vision | 1.1.0 | --- name: screen-vision description: > AI screen vision and desktop computer control skill for OpenClaw. Let your AI agent see the screen, understand UI elements, and autonomously perform mouse and keyboard operations (click, type, scroll, drag) via a screenshot-analyze-action loop. Cross-platform: Linux (headless serv... | [
{
"path": "config.example.json",
"content": "{\n \"vision\": {\n \"provider\": \"\",\n \"baseUrl\": \"\",\n \"apiKey\": \"\",\n \"model\": \"\"\n },\n \"safety\": {\n \"max_duration_min\": 5,\n \"max_actions\": 100,\n \"confirm_before\": [\"delete\", \"format\", \"rm\", \"sudo\", \"p... | suspicious | high | gpt-5.5 | The skill’s desktop-control purpose is clear, but it has unsafe defaults and weak safeguards for a tool that can view and operate a live computer. | clean | 0 | [] | clean | 0 | 0 | 0 | 67 | suspicious | 100 | CRITICAL | 20 | [
"AST5",
"E1",
"EA2",
"LP3",
"SDI-4",
"SQP-1",
"SQP-2",
"TT2"
] | [
"Dangerous Code Execution",
"Data Exfiltration",
"Excessive Agency",
"MCP Least Privilege",
"Data Flow"
] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:33:15.172000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
f2cd06be1caa03d17d60cd1269e4757b82a806b1d79c8db7060dd70b8362ee74 | yiguoguo/communication-coach | 1.0.3 | --- name: communication-coach description: "沟通教练。Challenge + 降维表达。Yi 在 EIZ 做 solution architect 经常要给各种人汇报、评审、谈判，试过不少沟通方法，这个是实战中总结出来的。" metadata: {"openclaw": {"emoji": "🥊"}} --- # 沟通教练 🥊 实战沟通训练。不是教你说话技巧，是帮你发现自己表达里的漏洞。 ## 两个核心能力 **Challenge 模式** 以一个挑剔的听众角度，对你的话发起追问。不是抬杠，是模拟真实场景里别人会怎么质疑你。 通常别人会问： - "前提是什么？" - "有数据吗？" -... | [] | clean | high | gpt-5-mini | An instruction-only communication-coach that matches its description, requests no env vars/binaries/install, and contains no code or unexpected endpoints. | clean | 0 | [] | clean | 0 | 0 | 0 | 66 | null | null | null | 0 | [] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:37:23.923000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
f47479b7e82470c3ab0195622eba38abecdffc6e183350d66e747095faf280e4 | besty0121/xiaohongshu-browser | 1.2.1 | --- name: xiaohongshu-browser description: Browse Xiaohongshu (小红书) and take screenshots of posts. Supports keyword search, post modal screenshots, and returns post links. Requires prior manual login. tags: - xiaohongshu - social-media - scraping - screenshot requires: bins: - python env: [] --- # Xiaohongshu Browser /... | [] | suspicious | medium | gpt-5.5 | The skill appears to automate Xiaohongshu browsing as advertised, but it stores reusable login/session data locally without clear security warnings or full disclosure. | suspicious | 1 | [
"suspicious.destructive_delete_command"
] | clean | 0 | 0 | 0 | 66 | suspicious | 52 | HIGH | 4 | [
"LP3",
"SDI-1",
"SQP-2"
] | [
"MCP Least Privilege"
] | {
"static": {
"status": "suspicious",
"verdict": "suspicious",
"reason_codes": [
"suspicious.destructive_delete_command"
],
"summary": "Detected: suspicious.destructive_delete_command",
"checked_at": "2026-05-10T06:28:57.680000Z",
"scanner_version": "v2.4.23",
"findings": [
... | train |
7446b04588bd511ac93855755749b1ef124bc1c2828843c4f80c116bf7a8731e | jason-aka-chen/quant-data-platform | 1.0.0 | --- name: quant-data-platform description: Comprehensive quantitative data platform for A-share market. Real-time quotes, historical data, alternative data (sentiment, news, fundamentals), factor data, and data quality monitoring. Essential infrastructure for quantitative trading. tags: - quant - data - trading - a-sha... | [] | suspicious | high | gpt-5.5 | This is not overt malware, but it is flagged for review because it presents randomly generated market data as real trading data. | clean | 0 | [] | clean | 0 | 0 | 0 | 62 | suspicious | 52 | HIGH | 4 | [
"SDI-1",
"SDI-4"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:21:01.085000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
983283bd168b055cfaaab2a4142c3a10b04071ed3ee52a5ef8497b5801b2242b | mhogan2013-9/ai-music-creator | 1.0.0 | --- name: ai-music-creator version: "1.0.0" displayName: "AI Music Creator — Generate Music for Your Videos" description: > Turn a 60-second travel montage clip into 1080p music-backed videos just by typing what you need. Whether it's adding AI-generated background music to videos or quick social content, drop your vid... | [] | clean | medium | gpt-5.5 | The skill is a disclosed cloud video/music generation connector, but users should be aware that it can create sessions and send media or prompts to NemoVideo when used. | clean | 0 | [] | null | null | null | null | null | clean | 20 | LOW | 2 | [
"SQP-1"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-05-21T12:34:24.352000Z",
"scanner_version": "v2.4.24",
"findings": []
},
"skillspector": {
"status": "clean",
"verdict": "SAFE",
"chec... | train |
22f362f0110a9f2d7bf4bdc450343de951b4e88b66690347c01e6e66d683664e | ashterry/asmr | 1.0.2 | --- name: asmr-music description: "Provide professional ASMR music recommendations and playback services to help users achieve sleep aid, heart rate regulation, and focus enhancement through scientifically designed audio. Use when user wants to (1) relax, sleep, or fall asleep, (2) regulate heart rate or relieve anxiet... | [] | clean | high | gpt-5.5 | This is a simple ASMR music recommendation skill with disclosed external playback links and an optional reminder feature. | clean | 0 | [] | clean | 0 | 0 | 0 | 57 | clean | 20 | LOW | 2 | [
"SQP-1",
"SQP-2"
] | [] | {
"static": {
"status": "clean",
"verdict": "clean",
"reason_codes": [],
"summary": "No suspicious patterns detected.",
"checked_at": "2026-04-30T05:18:22.058000Z",
"scanner_version": "v2.4.5",
"findings": []
},
"virustotal": {
"status": "clean",
"reason_codes": [],
"checke... | train |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.