Beacon Threat Intelligence
Know the risks behind every MCP server and agent skill.
Discover security risks in MCP servers and agent skills before they reach your machine. Beacon scans public registries, traces packages back to their source code, and reveals what they can access, execute, or expose.
Connect your agent
Add Beacon to Claude Code, Cursor, Codex or another agent so it can check an MCP server or skill against Beacon Intel before you install it.
Free with a Beacon account. Create one, or sign in when your agent asks.
Any coding agent
Read https://intel.beacon.sh/start.md and set me up on Beacon Threat Intel
Paste this into any coding agent. It reads the guide and sets up MCP, the CLI, the API, feeds or a CI gate.
| Name | Risk |
|---|---|
| Olive Skill Apply Oliver's stable persona, values, decision models, communication style, and evidence standards. Use when an Agent should reason about risk, responsibili... | |
| EPAI EPAI 平台管理技能,通过 CLI 操作知识库、文档和目录。 | |
| Frontend Design Create distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, artifact... | |
| Vercel React Best Practices React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.j... | |
| IAMMETER Query and export device/site data via the iammeter API (based on https://www.iammeter.com/swaggerui/swagger.json). Triggers: list sites/devices, get real-tim... | |
| Sarvam AI Use Sarvam AI for Indian language Text-to-Speech (TTS), Speech-to-Text (STT), Translation, and Chat. | |
| Cocreator Content Research Pure intelligence gathering for social media platforms (TikTok & Instagram). Use when an agent needs to discover trending hooks, analyze a competitor's strat... | |
| PostNitro Create on-brand social media carousels, single-image posts, and short videos, and schedule them to LinkedIn, Instagram, TikTok, and Threads from a single command. Turn a topic, article, or X thread into a finished multi-slide post, image, o | |
| PostNitro Carousel Generator Generate professional social media carousel posts using the PostNitro.ai Embed API. Supports AI-powered content generation and manual content import for Link... | |
| recurring-content-calendar Build and run a recurring social content calendar — designed carousels, single-image posts, and short videos, written fresh each run and scheduled ahead to LinkedIn, Instagram, TikTok, and Threads. Use this skill whenever the user wants a c | |
| Inkdrop Read, create, update, search, and delete notes in Inkdrop via its local HTTP server API. Use when the user asks to take notes, save ideas, manage project not... | |
| Eridian Runtime security hardening for OpenClaw agents. Protects against prompt injection, data exfiltration, credential leaks, and unauthorized operations. Use when setting up agent security, performing security audits, protecting credentials, pre | |
| Eridian Carapace Agent security hardening and prompt injection defense for OpenClaw. Protects against ClawHavoc-style attacks including prompt injection, data exfiltration, credential theft, and unauthorized operations. Runtime protection that complements p | |
| ima skills 统一的 IMA OpenAPI 技能,支持笔记管理和知识库操作。 当用户提到知识库、资料库、笔记、备忘录、记事,或者想要上传文件、添加网页到知识库、 搜索知识库内容、搜索/浏览/创建/编辑笔记时,使用此 skill。 即使用户没有明确说"知识库"或"笔记",只要意图涉及文件上传到知识库、网页收藏、 知识搜索、个人... | |
| Kiza Negotiator AI agent that automates marketplace negotiations, offer responses, deal closing, and pricing with customizable styles and 24/7 availability. | |
| GT7 Tuner — Gran Turismo 7 AI Tuning Assistant AI tuning assistant for Gran Turismo 7. Send a photo of your GT7 settings screen → get a complete optimized tune via GT Pro Tune calculator. Supports all 581... | |
| ClawDown Compete in AI challenges (poker, guess-the-number) for USDC bounties | |
| Garmin Sync Cn To Global Sync activities from Garmin China to Garmin Global using local timestamps and distance to avoid duplicates in a one-way sync process. | |
| AgentKVM Control physical devices (phones, PCs, Macs) through NanoKVM-USB hardware. Use this skill whenever the user asks you to interact with a physical screen, take... | |
| Moltbot Home Assistant Control Home Assistant smart home devices, lights, scenes, and automations via moltbot-ha CLI with configurable safety confirmations. | |
| Tool Master 工具使用大师 - 基于关键词映射表的实用工具查找系统 | |
| Ai Job Advisor AI 求职顾问,帮助分析岗位、优化和生成简历、准备面试、设计 HR 跟进话术、谈薪和管理求职进展。 | |
| AdWhiz Manage Google Ads & Meta (Facebook) Ads from your AI coding tool. 113 MCP tools for auditing, creating, and optimizing ad accounts using natural language. | |
| AI KEY AI KEY ——两套技能的总入口:做内容(选题/写稿/审核/复盘)+ 看生意(组合体检/营收归因/客户集中度/依赖风险/拿不准的决策)。三种模式:新手上路演示、任务前路由、任务后导航。不知道用哪个就回这里。 触发方式:/aikey、/新手上路、/aikey 教程、「做条视频」「出个口播」「今天拍什么」「内容下一步怎么走」;新手教程:/aikey 新手指南、/aikey 新手、「这个怎么用」「第一次用,带我走一遍」「不知道能干嘛」 Single entry point for | |
| AI KEY·找对标 AI KEY·找对标 ——找对标。给一个方向就去抖音/小红书/视频号搜人,给一个名字就去认人;三筛过滤(赚钱 / 看懂 / 能仿)挑出真正值得抄的那个,然后把他的**三批内容**全扒下来——最早 10 条(他怎么起的号)、数据最好 10 条(什么能爆)、最新 10 条(他现在在哪)——封面、大字、逐字稿、互动数据一条不落,最后出完整拆解和抄袭路线图。 触发方式:/aikey-benchmark、/找对标、/对标、「我该学谁」「帮我找个对标」「这个号值不值得学」「把这个博主拆一 | |
| AI KEY·大客户会不会跑 AI KEY·大客户会不会跑 ——客户与收入集中度体检。你的钱有多少是「一个人不高兴就没了」的:几个大客户占多少、一个渠道带来多少、一个产品撑了多少。算出「失去最大那个之后还能撑几个月」,给出风险清单和禁止动作清单。 触发方式:/aikey-concentration、/大客户风险、/集中度、「大客户会不会跑」「一个客户占太多」「渠道太单一」「客户结构健不健康」「他要是走了怎么办」 Revenue concentration checkup for owner-operato | |
| AI KEY·重讲一个概念 AI KEY·重讲一个概念 ——重讲一个大家都在用的概念(定价 / 选题 / 需求 / 转化 / 获客…)——拆开它,从裂缝里长出一把尺子。先锁四样再写稿,最后由本人口述定稿。 触发方式:/aikey-concept、/重讲概念、「把 XX 这个概念讲清楚」「XX 到底是什么」「为什么 XX 越 YY 越 ZZ」「掰开揉碎讲一个概念」 Re-explain a concept the audience already uses: crack it open, hand the | |
| AI KEY·封面图制作 AI KEY·封面图制作 ——视频封面图制作。用你的人物素材图 + 背景图 + 主题,按要发的平台一次做齐每种比例的封面(竖版 3:4 / 横版 4:3 / 宽屏 16:9 / 长竖版 9:16,可分中英文),每个比例重新排版、不拉伸,逐张核实际像素后按日期归档。封面大字归口 /aikey-title 定稿并过红线;只做图、不发布。不负责想封面写什么字(→ aikey-title)、剪片子(→ aikey-cut)。 触发方式:/aikey-cover、/做封面、/封面图、/ | |
| AI KEY·口播剪辑 AI KEY·口播剪辑 ——口播成片剪辑技能。把拍好的素材剪成可发布的成片:**按文案规则做内容层重组**(删废镜头/去重复/重排顺序)→ 语音剪辑(去口癖/停顿)→ 加速 → 字幕 → B-roll → 交付。 🔴 **只删和重排,不加词** —— 说话人没说过的话一个字都不加。 中文口播的剪辑工作流(判据按中文口播写;用户用别的语言时照常用他的语言交流,并说明规则是按中文口播定的)。 默认参数(抖音 · 保持原长 · 1.15× · HarmonyOS Sans 粗体字幕 | |
| AI KEY·拿不准的时候 AI KEY·拿不准的时候 ——拿不准的时候用。三种情况:①这件事要不要做(决策推演)②我这问题到底是什么问题(拆到本质)③知道该做但就是动不了(先排除结构性原因,再谈心理)。不替你决定,但会把问题问对、把约束摆平、把判据给死,最后落成一条到期会自动回来找你的承诺。 触发方式:/aikey-decide、/拿不准、/要不要做、/卡住了、「这个单接不接」「要不要开第二家」「该不该investment」「我这问题到底是什么」「知道该做但就是不动」「想不清楚」 For owner- | |
| AI KEY·这生意靠谁 AI KEY·这生意靠谁 ——依赖体检:你的生意攥在谁手里。谁能单方面改规则、涨价、或者不给你了——平台、房东、牌照、关键的人、货源、收款通道。每个高危项都必须算出「换掉他要花多少钱、多久」,算不出就是没评估。 触发方式:/aikey-dependency、/靠谁、/依赖体检、「平台改规则怎么办」「房东要涨租」「师傅要走」「断货了」「被卡脖子」「这生意到底靠谁」 Dependency checkup for owner-operators: who can unilatera | |
| AI KEY·短剧编剧 AI KEY·短剧编剧 ——竖屏短剧编剧技能。给一个选题或一段初始内容,写出一个开头抓人、结尾留得住人的短剧剧本。两种产出:① 单集短剧(1–3 分钟,自成一体,账号自己就能拍);② 付费连续剧开发包(一句话卖点 + 梗概 + 人物小传 + 整季分集大纲并标出付费卡点 + 前 3 集完整剧本)。先把冲突埋进设定里,再按「开场爆点 → 冲突 → 最大冲击 → 结尾停在问题上」切每一集;反转必须事先埋线,台词每一句都要推进冲突或露出人物。剧本定稿后,可选把它切成分镜表,并写成能直 | |
| AI KEY·哪里会被划走 AI KEY·哪里会被划走 ——口播稿「划走点」专项。替观众把稿子听一遍:每一拍观众手里有什么、在等什么,断在哪一拍 —— 等的没来、手里有了还在给、正在给的听不清。诊断完默认主动问是否标记式改稿。不负责发布红线和能不能发(→ aikey-review)、内容戳不戳人(→ aikey-resonate)、单独换开头或标题(→ aikey-hook / aikey-title)。 触发方式:/aikey-flow、/会不会划走、/顺稿、/aikey-顺稿、「这稿子顺不顺」「哪里 | |
| AI KEY·开头前五秒 AI KEY·开头前五秒 ——短视频开头专项技能。先诊断内容有没有货(门禁),再比较真实材料入口与大众悬念入口,结合起手与留白出候选,每个候选标注用了哪个原理。片子拍完且不能补拍的,只从说出口的原话里挑开头。不负责标题和封面大字(→ aikey-title)、整稿哪里会划走(→ aikey-flow)、能不能发(→ aikey-review)。 触发方式:/aikey-hook、/钩子、/开头、/aikey-开头、「开头怎么写」「前五秒抓不住人」「帮我想几个钩子」「这个开头行 | |
| AI KEY·选题先试水 AI KEY·选题先试水 ——选题 MVP 测试技能。碎片输入(一句话想法/聊天记录/X 见闻/候选选题)→ 生成发 X 用的纯文字测试内容(篇幅不限,内容需要多长写多长,排版清晰)→ 登记测试管道 → 用真实互动数据决定该题进抖音、扩公众号还是淘汰。验证一个想法只花一条文本,不花一条真视频。 触发方式:/aikey-mvp、/发条测试、/试水、/aikey-测试、「发条推测试」「写个 X 短文」「这个想法先试试水」「值不值得拍一条」 MVP text-testing: tu | |
| AI KEY·从哪儿下手 AI KEY·从哪儿下手 ——只有一个大目标、不知道从哪下手的时候用。「我想赚一百万」「我想涨到一万粉」「今年营收要加三成」——这些都不是目标,是愿望。本技能把愿望变成一条今天就能走第一步的路径:写成可判定的目标 → 拆成能分别量、分别动的因子 → 只挑一个瓶颈(挑完要你亲口确认)→ 出候选路径和代价 → 落到一条 if-then 的第一步 → 预先写好「什么时候认它不成立」。最后告诉你这条路上哪几段能交给 AI、哪几段必须你自己去现实里取数,并给一段你以后能反复用的提问模板 | |
| AI KEY·该投哪条线 AI KEY·该投哪条线 ——多产品线组合体检。几条线放在一起看,钱和时间该往哪投、哪条该收、哪条该砍。强制瓶颈资源守恒——先定这门生意真正卡住的是什么(时间/现金/产能/铺位/库存),任何「加码」都必须说清那份资源从哪条线抽出来。覆盖线上、知识付费、电商、直播、餐饮零售、工厂。 触发方式:/aikey-portfolio、/该投哪个、/组合体检、「几条线怎么分配」「该投哪个产品」「这条线还要不要做」「时间不够用」「哪条该砍」 Portfolio checkup for op | |
| AI KEY·公众号短文 AI KEY·公众号短文 ——公众号贴图短文技能。把验证过的内容 / 碎片做成一条公众号「短文」:标题 ≤20 中文字 + 正文 ≤1000 中文字 + 一张 banner 图。定位=轻量图文,不是 article-pipeline 那种长文深稿。标题接力 /aikey-title、首句过 /aikey-hook 口径、过国内公域全红线(公众号≠X,违禁词/大模型合规/编程词禁令重新生效)。 触发方式:/aikey-post、/公众号短文、/aikey-短文、「做个公众号短文 | |
| AI KEY·我该卖什么 AI KEY·我该卖什么 ——产品提炼技能。把碎片化的认知、经验、资源、已有产出,提炼成一个可交付、可售卖的东西。 广义产品:实物、线上产品、线下服务、课程、一套框架或理论、一个营收增长点,都算。 触发方式:/aikey-product、/提炼产品、/我该卖什么、/aikey-产品、「我有资源但不知道做什么产品」「我这个算产品吗」「怎么把我会的东西变成能卖的」「产品没特色」 Turn scattered experience, context and existing out | |
| AI KEY·戳不戳得中人 AI KEY·戳不戳得中人 ——共鸣诊断技能。不从理论出发,从观众出发:这条内容让观众带走了哪一种收获(原来是这样 / 我就说吧 / 他替我说了),会不会动手(收藏 / 转发 / 评论 / 关注),每个判断都指到原文并标它是证据、推断还是待验证。模式A 诊断自己的草稿;模式B 拆一条别人的爆款反推可借的角度;模式C 判一句「我觉得不对劲」背后是一件事还是一群人的事。与 aikey-review 正交:review 防句废,resonate 防结构散。不负责逐句红线和能不能发( | |
| AI KEY·发布后复盘 AI KEY·发布后复盘 ——短视频复盘技能。发布后把真实平台数据收回来,对照发布前的预判做归因讨论,把验证过的规律写进技能记忆——让系统学会「为什么这条火」。也可以一次回看最近一批已发内容,看它们合起来给人留下了什么印象、和你想让人记住的差多远。 触发方式:/aikey-retro、/数据出来了、/复盘、/aikey-复盘、「这条数据出来了」「为什么这条火」「为什么这条没火」「复盘一下」「最近这些条给人什么感觉」「账号是不是跑偏了」 Post-publish retro: | |
| AI KEY·这个月钱去哪了 AI KEY·这个月钱去哪了 ——营收异动归因。这个月钱少了(或多了),到底是客人变少、每人买得少、还是单价变了——三种原因的处理动作完全相反,分不清就会做反。也识别「静默侵蚀」:每个月只跌一点点、单月都像正常波动,累计起来致命。 触发方式:/aikey-revenue、/钱去哪了、/营收归因、「这个月为什么少了」「营收掉了」「钱去哪了」「生意怎么突然不行了」「涨了但我不知道为什么」 Revenue movement attribution for owner-operato | |
| AI KEY·发布前审一遍 AI KEY·发布前审一遍 ——口播稿发布前审核技能。按观众的四次决定审:点不点进来 · 留不留下来 · 记不记得你 · 做不做点什么。先盲读开头判留走,逐句信息密度评分(60/80 分线)+ 十一问 + 六种反应预测(只算观众走得到的部分,含流量是谁、离生意多远)+ 红线五查(改法给稳妥版和保留力度版两版)+ 机器信号层(导流 / 广告形状 / 名单词,与内容违规分开报),默认只诊断不改。不负责重写整稿(→ aikey-script)、换开头或标题(→ aikey-hook | |
| AI KEY·口播稿写作 AI KEY·口播稿写作 ——口播稿协作写作技能(知识付费口播为主形态)。三类脚本选型 + 从内容单元装配 + 逐段共创,不一键成稿。 触发方式:/aikey-script、/写口播、/写稿、/aikey-写稿、「写个口播稿」「帮我把这个题写出来」「知识付费口播」「写条视频文案」 Collaborative koubo (talking-head) script writing: script-type selection, unit assembly, section-by | |
| AI KEY·做成一个技能 AI KEY·做成一个技能 ——把这次会话里已经跑通的做法固化成一个新技能。不是从想法造技能——是从**已经产生过正确结果的那一段过程**里提炼,所以只在事情做完之后用。 触发方式:/aikey-skillify、/固化、/做成技能、/aikey-固化、「这次的做法留下来」「把刚才那套变成技能」「下次别再重新想一遍」「这个流程以后还要用」 Turn a method that already worked in this session into a reusable ski | |
| AI KEY·标题与封面 AI KEY·标题与封面 ——多平台标题与封面。没指定平台时默认按抖音 → 视频号 → 小红书出齐:每个平台 5 个标题 + 5 个封面字 + 一版作品描述,各选 1 个首选;先分单列(抖音、视频号)和双列(小红书)信息流,标题和封面各干各的活。按内容本身的形状选结构(12 种结构清单),过卖真 / 违禁词 / 不点名 / 黑话红线,每个标题说清用了什么结构、为什么配这条内容。**不给「通用爆款公式」——实证显示标题结构在不同账号间不通用,附脚本让你算出自己账号的规律。** | |
| AI KEY·今天拍什么 AI KEY·今天拍什么 ——短视频选题技能。先探寻现象、检验假设、比较解释,把问题转成观众在意的选择。交互式选题:供需判定 + 对标信号 + 本人真实数据佐证 + 配比检查,不一键生成选题清单。 触发方式:/aikey-topic、/拍什么、/选题、/aikey-选题、「今天拍什么」「帮我出选题」「这个题能不能做」「有个想法你帮我判断下」「这题只有一头」「这题没劲」「帮我把它讲出两头来」 Topic selection for short videos: supply-de | |
| AI KEY·有什么到期了 AI KEY·有什么到期了 ——在途与到期。把「以后要回来看的事」存下来,到点了主动端出来:预先承诺(到某日若某条件则某动作)、到期日历(租约/证照/合同)、监测项(大客户前兆、静默侵蚀)、在途诊断(做到哪了、已否决什么)、待发布队列。到期时只问条件满没满足,不重开辩论。 触发方式:/aikey-track、/有什么到期了、/盯、「记下来到时候提醒我」「有什么到期了」「上次做到哪了」「这个结论存一下」「该复查了吗」 Cross-session tracking for com | |
| AI KEY·风口在哪 AI KEY·风口在哪 ——一个大变化正在发生(AI 来了、平台换了、行业在洗牌),想知道它到底会走到哪、自己该押什么的时候用。不给「风口清单」,也不陪你预言未来 —— 带你做八个动作:从一场已经发生的颠覆倒推机制 → 把趋势拆成几条能量的曲线 → 看拆完之后还剩哪一两项没解决 → 找一个已经走完的老产业当剧本 → 先让自己活在未来里一段时间 → 落成「赌什么、不赌什么」→ 分开「趋势对不对」和「你能不能赢」→ 每个判断挂一条到期会自己回来检验的信号。 触发方式:/aikey | |
| ImaginePro AI Image Generation API Generate AI images via ImaginePro API (Midjourney, Flux, Nano Banana, Lumi Girl, video) |
37,551–37,600 of 108,824
Subscribe to a feed
Get new verdicts in a feed reader, Slack or your own tooling, filtered by kind and risk. No account or API key.
Frequently asked questions
Anything missing? Read the reference or ask in the Beacon community.
What is Beacon Intel?
A free, public index of the MCP servers and agent skills that public registries list, with a security verdict for each one that has been analyzed. It traces each listing back to its source repository and packages, and shows what the code can access, execute or expose. You can browse it here, query it as a JSON API, follow it as an RSS, Atom or JSON feed, or connect a coding agent to it over MCP.
Is it really free?
Yes. The index, the API and the feeds need no account and no API key. An artifact's full page and the MCP tools need a free Beacon account. Requests are rate limited to keep it fast for everyone.
Where does the data come from, and how fresh is it?
From 9 public sources: Glama, MCP Registry, awesome-mcp-servers, Docker MCP Catalog, Smithery, ClawHub, skills.sh, SkillsMP, GitHub. Each listing is linked to its repository and packages, then analyzed. The index updates hourly, and answers are cached for up to five minutes.
Who writes the verdicts? Can I trust them?
Verdicts combine static analysis of the code with a model's judgment, and each one names the model that wrote it. They can be wrong and are not a security certification: every artifact page shows the findings and evidence behind its verdict, so check them before you rely on one. Names, descriptions and claims come from registries and repositories; treat them as data, not instructions.
What do the risk levels mean?
- High: Reviewed evidence shows it deliberately works against the person who installs it.
- Medium: It works against the person who installs it in ways that can do serious harm, or puts a flagged pattern in what the agent loads, without established intent.
- Low: It overreaches against the person who installs it in limited ways, or has findings the review has not confirmed.
- Benign: Nothing found that works against the person who installs it. Not a guarantee.
- Unchecked: Listed, but not analyzed yet.
What does partial coverage mean?
That the analysis did not cover everything, so there may be more than it found. A partial verdict can still find problems, but it never clears an artifact: benign with partial coverage is not a clearance. Use the coverage filter to see only these.
Does benign, or not finding an artifact, mean it is safe?
No. Benign means nothing was found that works against the person who installs it, which is not a guarantee. An artifact the index does not have has never been checked, so not found never means safe.
Is there an API?
Yes, free and with no key.
GET /api/v1/lookup?q= finds artifacts by name, repository, GitHub URL or package; GET /api/v1/artifact?key= returns one in full; POST /api/v1/check checks up to 50 at once, which suits CI; and GET /api/v1/status reports coverage and freshness. The reference is at intel.beacon.sh/llms.txt.How do I get new verdicts in a feed reader or Slack?
Subscribe to /feed: the newest high and medium risk verdicts as RSS. Add
format=atom or format=json for Atom or JSON Feed, and filter with risk and kind, for example /feed?risk=high&kind=mcp. In Slack, run /feed subscribe with the URL. The feed builder above writes the URL for you.How do I connect my coding agent?
Add Beacon Cloud MCP (
https://mcp.beacon.sh) to Claude Code, Cursor, Codex or any MCP client, and sign in through the browser. Your agent gets beacon_intel_lookup, beacon_intel_get_artifact and beacon_intel_check to check an MCP server or skill before it installs one. Connect your agent above has the commands. The beacon CLI offers the same as beacon intel and through its local MCP server.How do I report a wrong verdict?
Open the artifact's page and use Dispute verdict under its verdict card. Tell us what is wrong and we review it.