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 |
|---|---|
| SEO Optimizer Provides keyword research, meta tag analysis, content SEO scoring, and meta description generation to improve website search rankings. | |
| Social Media Manager AI-powered assistant for creating, scheduling, and analyzing social media content with platform-specific optimization and hashtag research. | |
| Social Media Manager AI assistant for creating, scheduling, and optimizing social media content with hashtag research, multi-platform posting, and engagement analytics. | |
| Surf Your AI agent's crypto brain. One skill, 83+ commands across 14 data domains — real-time prices, wallets, social intelligence, DeFi, on-chain SQL, prediction... | |
| Surf AI Crypto Skill Your AI agent's crypto brain. One skill, 83+ commands across 14 data domains — real-time prices, wallets, social intelligence, DeFi, on-chain SQL, prediction... | |
| testcase-generator-skill 功能测试用例智能生成工具。用于根据需求文档、接口定义、数据库表结构、现有用例参考等输入， 自动生成标准格式的测试用例文档（Excel格式）。适用于：(1) 解读需求文档生成测试点 (2) 基于接口定义生成API测试用例 (3) 根据数据库表设计验证数据测试 (4) 参考现有用例生成 相似模块用例 (5) 批量生成... | |
| Memory Management (PARA) Manage and maintain memory system using the Root + PARA hybrid architecture. Use when performing memory maintenance, distilling daily logs, and updating core... | |
| SwarmMemo Read and post on SwarmMemo, the public square where AI agents meet, work and keep their word, and find paid tasks other agents posted. Use when the user wants to talk with other AI agents, find work an agent can be paid for, post a question | |
| Voice Ai Integration Integrate Shengwang products: ConvoAI voice agents, RTC audio/video, RTM messaging, Cloud Recording, and token generation. Use when the user mentions Shengwa... | |
| Public Opinion Scanner 品牌/个人舆情扫描器。根据公司名或人名，从百度、搜狗、微博、知乎、财经网站等来源搜索近1年的公开信息， 逐条标注正面/负面/中性，保留原始链接，生成结构化 MD 舆情分析报告。 触发词：舆情扫描、品牌声誉分析、公司舆情、个人舆情、公众形象分析、搜索XX舆情 | |
| Commit code safe and nice Smart git commit with remote sync, amend intelligence, and conventional commits. Use when the user asks to commit changes, stage and commit, "/commit", save... | |
| Publish Wiki Pages Publish markdown content to Wiki.js with proper formatting and metadata. Use when user wants to create or update wiki pages, convert notes/articles to wiki f... | |
| Query Payful Account Query Payful account information including balance, transactions, and account details. Use when the user needs to check their Payful account status, view bal... | |
| Web Site or Domain Name Basic Information Scanner Comprehensive website analysis and scanning tool. Analyzes IP (IPv4/IPv6), DNS records, WHOIS data, website content (including Schema.org JSON-LD, robots.txt... | |
| Blog with Wordpress Publish articles to WordPress blogs via REST API. Handles post creation, category/tag management, and SEO-friendly English slug generation. Use when user ask... | |
| Memory Strategy 当用户需要管理对话记忆时应使用此skill。 触发条件包括： - 用户说"记下来"、"记住这个"、"别忘了"、"永久保存"、"这是一个重点" - 用户查询历史信息："之前是怎么做的"、"查找关于...的记录"、"回忆一下..." - 会话结束需要自动整理归档 - 需要评估信息重要性并决定存储位置 - 需要更新记忆... | |
| The skill for chess player Affiche les stats Chess.com d'un joueur (rapid/blitz/bullet + puzzles) via l'API publique. | |
| Wiz by TradingWizard Use Wiz for current market prices, macro mood, Market Track, movers, global TradingWizard bots, ranked opportunities, Bot Fund performance, and optional private account tools through OAuth. | |
| libtv-skill agent-im 会话技能 - 通过 liblib.tv 的 AI 能力生成和编辑图片/视频，兼容 Gemini 文生图、Sora/Veo/Grok/豆包/Vidu 视频直连。 覆盖：文生图、文生视频、图生视频、视频编辑、风格迁移、视频续写、MV生成、短剧生成、分镜设计。 触发词：画、生成、做动画、改镜头、换风格... | |
| flooods Complete knowledge base for Floods' business model, product mechanics, pricing, verification, attribution, brand safety, and competitive positioning. Use thi... | |
| heursistic Master reference for 88 psychological heuristics — the complete leverage list for sales, copywriting, offer design, cold outreach, objection handling, and pe... | |
| market sales skills Market context, strategic thesis, case studies, proof points, sales pipeline, and B2B communication guidelines for Floods. Use this skill when working on pit... | |
| ooutreachskill Complete outreach operating system for Floods: contact tiering, personalized DM writing, prospect research, hook crafting, follow-up sequences, and cold outr... | |
| fk-boss-greeting Boss直聘打招呼自动化工具。用于搜索职位并向HR主动打招呼。当用户要求在Boss直聘(老板直聘)上发打招呼、投递简历、搜索职位并沟通时使用。支持：搜索Java/架构师/技术经理等职位，自动过滤黑名单公司(阿里、百度、腾讯、字节、京东等)和无关岗位(算法、AI、测试、产品经理等)，逐个排队打招呼(间隔1秒)，记录... | |
| Bijie Express Tracking Fast package tracking and logistics timeline lookup for 2000+ couriers with BijieServ. Use for 查快递, 快递查询, 查物流, 物流轨迹, 快递到哪了, 包裹追踪, track package, where is my... | |
| 公告速览·A股 A股公告速览：输入一家公司名／六位代码或一份自选清单，产出当日（或指定日期窗内）公告的逐条要点清单——每条带公告标题＋披露日期＋关键数逐字，同类事项自动串成前后两条对照，缺的点名「未检索到」。评级、目标价、走势判断一律 [待人工]，不预测、不代下单。Do NOT use for: 投资建议或买卖决策、未检索到的公告编造或补全、收益与风险承诺。Triggers: 今天有什么公告 / 当日公告 / 公告清单 / 公告速览 / 这周发了什么 / 自选清单公告； daily anno | |
| 债券条款卡 境内债券条款卡：输入一份债券募集说明书或发行方案，产出八项逐字速读——发行人与债项名称、发行规模与期限、票面利率与付息方式、发行价格与计息方式、担保与增信安排、赎回与回售条款、信息披露与受托管理、违约与交叉违约。每项挂原句＋锚，缺项点名「未检索到」；规模与利率逐字，推定语不算值；评级符号必须同句带锚。债项估值、收益率测算、评级与目标价一律 [待人工]。Do NOT use for: 投资建议与买卖决策、估值与收益率测算、代客认购、收益与风险承诺。Triggers: 这只债的条 | |
| 日历·催化剂 给你持仓列表 + 时间窗，产出按日期排的盯盘日历：每条事件的日期要么已有披露在场且含明确 YYYY-MM-DD，要么 statute 原文推导且适用条件在锚内可判；未披露的推测事件、任何日期算术都进不了表。机检对评级词与褒贬零容忍（词族以机检现值为准，本文所举为常见例）。适合：财报季前置、组合盯盘节奏、投研会材料。不适合：涨跌与择时建议、事件影响判断（本件只陈述不判断，也没有留白判断位）。豁免只有两条口径：类型列整列不扫、官方披露词组按词形豁免。 Do NOT use for | |
| 民事二审翻案工程·论断逐条带出处 民事一审败诉后的翻案工程：还原法官推理链、定位争点被框错的层；证据逐字比对截断/省略/割裂；请求权铺主攻/备位/兜底阶梯；事实台账三级出处；文书包同日递交与自行诉讼操作。不提供法律意见、不替代律师，条号与费用标准递交前核对。脚本与流程离线自包含。Triggers: 一审败诉上诉 / 二审上诉状 / 判决书拆解 / 为什么败诉 / 上诉期与继续保全 / 自行诉讼； lost at first instance / civil appeal / appeal deadline / | |
| 合同逐条要点与需盯条款清单·逐条挂原文 合同逐条拆解起草：输入是一份合同正文（本地文件或解析通道取回的正文），产出四节清单——一句话概览（标的、期限、价款，只写文里有的）、逐条要点（每条一句原文摘，挂 [S#] 与页码或行号）、需盯条款（单方变更／自动续约／排他／违约金／争议解决地／数据与保密／责任上限七类逐类点名，没有的写未见到）、来源与解析日。机检 check_clauses.py 五道：形制与条数带／锚与定位／引文逐字回原文／七类点名／法律意见与效力断言零命中，零网络。适合签前自查。Do NOT use fo | |
| 财报深度点评 给你一家公司 + 报告期，产出八节点评初稿：四期差异、科目勾稽、指引与兑现对照、盈利质量分桶、预测脚手架。适合：财报复盘、投研底稿、晨会；不适合：出具评级、监管申报、无公开披露渠道。四道机检的如实范围：①AI 声明与 [待人工] 就位；②散文数字行须有证据级标注，逐条拦+95%（哪些行算「数字行」以包内 check_note.py 两条正则现值为准，本文举例非穷尽；列表行与索引节按设计豁免）；③账本数值对象带basis字段；④合规词表最小集。--sources 仅给出时校验 | |
| 一条链接→一条带出处的笔记 链接→笔记起草：输入是一枚 URL（正文走解析通道取回本地那一份）或一份本地文件，产出是一条可直接进笔记库的 Markdown 笔记——标题、要点 3–7 条（每条一个 [S#] 锚）、至多一枚原话引用（逐字，带页码或行号）、口径与未取到、来源行（带 文件:行号 或 URL 与检索日）。机检 check_note.py 四道：形制／锚与定位／引用是否逐字回原文／缺口是否点名，零网络。适合读完一篇先落一条能回查的笔记。Do NOT use for: 整篇换格式、多源互校速览、建 | |
| 中文多源研究速览·逐条挂源 多源研究速览起草：输入是一批中文来源（网页／公告／报道／解析件）或一个待查问题，产出是四节速览——一句话结论、逐条发现（每条一个 [S#] 锚）、未取到与口径、来源（逐条带 URL 与检索日）。机检 check_brief.py 四道：形制／锚／判断词／未取到点名，零网络。适合投前初筛、行业快讯、竞品多点对照、政策多源核对。Do NOT use for: 评级、目标价、买卖与时机建议（一律 [待人工]）、单一来源的转述（那只是摘要）、逐数核口径的财报深度点评（走姊妹件那一路） | |
| 简历 × JD 双向核对·逐条挂原话 简历 × JD 双向核对起草：输入是两份文档（本地文件或解析通道取回的正文），产出四节报告——一句话结论（只说匹配到哪一层）、逐条对齐（每条要求一句、证据一句，各挂 [S#]，两句都须逐字）、缺口两栏（JD 有要求而简历无证据的、简历有内容而 JD 不关心的）、来源（文件:行号 或 URL ＋ 解析日）。机检 check_fit.py 五道：形制与条数带／双锚与定位／两侧逐字回原文／缺口点名／录用与排序断言零命中，零网络。适合岗位对照、转岗材料整理。Do NOT use fo | |
| 图表规格 表→图起草：输入是一张表（本地文件，或解析通道从扫描件／云盘 PDF／照片里取回的正文），产出四节——图的规格（图型／横轴／纵轴／序列，由人选不代选）、读表要点（每条一句、每个数配列头逐字回那一格、挂 [S#] 与页码或行号）、图的限度与未取到（缺格与空值／合并格与跨列表头／单位与口径／表外推与未读到，逐类点名）、来源与解析日。机检五道：形制与条数带／锚与定位／数逐字回那一格（全篇书名号都回表）／四类限度点名／投资与评级断言零命中，零网络、不代人算数。Do NOT use f | |
| 竞品决策简报·逐条带出处 给你主体 + 竞品清单，产出简报：结论先行（一页以内、含可执行建议），对比表每格对应证据表单元格——fact 直接可证 / inference 由多事实推导 / opinion 无可引用来源就标 opinion、置信度宁可空着不填「高」。流程硬规矩：证据表没填的维度不写进简报；结论性主张必须先进表；简报末尾列行数、置信度分布与完全没覆盖的维度。适合：尽调背景、会前准备、策略讨论。不适合：法律合规结论、非公开数据推断。 Do NOT use for: legal conclus | |
| 财报引文逐字核对 用 Cue Omni Reader 读公司自己的公开披露，找信用恶化线索，每句引文都对回原文和页码 | |
| 公开资料变化与证据简报 从报告和公告中回答业务问题，展开原文查证，加入新资料继续跟踪变化 | |
| 公开信息预尽调清单 给你一家公司 + 基准日 + 回看窗口 + 用途档位（投资/授信/并购），还你一张固定七列的风险事实表：每条都回指具体披露（公告号、函件号、裁判文书号或 URL），外加九类覆盖率对账表，写清每类查没查到、没查到是没工具还是真零条。不适合：投资意见、法律意见、批量多主体、公开面之外的深查（会标注需商用库）。适合：投前初筛、授信背景核查、并购线索起点。八道机检管形制：无锚、无日期、归属不清的条目进不了表，「无风险」「可投」类措辞出现即拦，账、流水、快照三者互相对账；但锚内容是否本 | |
| 公告事件速读卡·关键数逐字带出处 A股公告事件速读卡：把一份临时公告（回购/增减持/业绩预告/重大合同/处罚问询等）变成可核对的五段卡——逐字关键数+出处、影响链候选、同类对照、待核清单。评级与目标价一律[待人工]，不预测走势。适合「刚出了公告先看什么」「这份公告讲什么」。Do NOT use for: 投资建议或买卖决策、未检索到的数据编造、收益承诺。Triggers: 公告速读 / 公告解读 / 这份公告 / 回购公告 / 增减持 / 业绩预告 / 处罚 / 问询函 / 刚出公告； announcemen | |
| 企业风险扫描·三源可查不打分 企业风险一键扫描（数据腿实测版）：给你公司名或代码，产出一页扫描卡——行政处罚与监管措施逐条（带决定文书号/日期/机关）、对外担保要点、全球制裁执法面（涉跨境主体时）；每条挂官方出处，检索未达的面**明写未达不补空白**。适合「这家公司有什么处罚」「对外担保扫一遍」「出海实体制裁风险查查」等场景。Do NOT use for: 整体风险评级或违约概率、涉诉与股权链推断（通道当前未达，本件不装）、投资建议。Triggers: 企业风险扫描 / 处罚记录 / 监管措施 / 对外担 | |
| 投决备忘录起草·逐条挂源 投决备忘录草稿生成：输入是一家公司名／六位代码或你手上的材料，产出七节备忘录——摘要、事实与数字、业务要点、风险与反方、缺数与未决、人工复核、来源索引；每个数字带披露或监管源＋字符/页定位，取不到的行直接写「未取到」，正反两方各至少一条否则机检拦下。适合投前材料整理、内部讨论稿、需要「每条数都能回原文核」的场合。Do NOT use for: 评级、目标价、买卖建议（一律 `[待人工]`）、最终决议（本件只起草，用户负终审）、估值测算（无行情与持仓数据腿，只做适用性声明）、整 | |
| 保险条款卡 境内保险条款卡：输入一份保险条款与投保须知，产出八项逐字速读——险种与保险责任、责任免除、保险金额与免赔额、等待期与观察期、保险期间与缴费期间、理赔申请材料与时限、犹豫期与退保规则、如实告知与合同解除。每项挂原句＋锚，缺项点名「未检索到」；金额与期间逐字，推定语不算值。赔付测算、保费与保额假设、能否获赔的结论一律 [待人工]。Do NOT use for: 投保建议与产品比价、理赔代办、条款编造或补全、收益与风险承诺。Triggers: 这份条款怎么赔 / 免赔额多少 / 等 | |
| cue-journal-draft 给你栏目 + 本期素材（或先做公开检索），产出可编辑成稿：提纲、成稿、引证逐条回源，后接版面与编辑定位两道脚本门禁（查定位落地、字数、读者动作、句长、自填 taboo 词表与术语密度），终了 QA 清单人工逐项核（引证回源、极限词在清单内）。仿写参考样式只学语气、不编事实。适合：月刊/季刊内刊、客户通讯、投资者信、ESG 特刊。不适合：替代编辑签发（门禁给清单、拍板在用户）、编造客户案例、承诺刊发效果。 Do NOT use for: editorial sign-off, | |
| 财报引文逐字核对 用 Cue Omni Reader 读公司自己的公开披露，找信用恶化线索，每句引文都对回原文和页码 | |
| 纪要·录音成稿 企业会议纪要：给我会议录音（本地文件，以通道返回为准）与可选材料，产出五段纪要——基本信息、逐字带时间锚、话题分段（归类是笔记法非原文）、决策与待办每条挂原话锚、待核清单。Do NOT use for: 实时转写、单件超 256 MiB 的整场录音直投（闸按字节不按时长；单件 ≤256MiB 为已实证边界）、替你把纪要改写成「更漂亮」的版本。适合「这段录音整理成纪要」「会是哪几件事谁负责」「纪要逐条对回原话」等场景。Triggers: 会议纪要 / 录音整理 / 谁负责什么 | |
| 任意文档·建逐句可回查知识库 从报告和公告中回答业务问题，展开原文查证，加入新资料继续跟踪变化 | |
| 表格还原 把 Omni 解析结果重建为 Word / HTML / PPT：页锚点分页、真标题+目录、表格脚注溯源、多源证据附录。Do NOT use for: 像素级版面复刻、无 omni-reader 的解析层（--md 直出除外）；Triggers: Omni 转 Word/HTML/PPT / 保真重建 / 多源证据整合 / 扫描件转 Word / 公众号排版 / 汇报 deck; omni to docx / parse to deck。开关与制式：--profile 公文/ | |
| 论文研读卡·观点可追句、引文可回源 学术论文研读包：给你一篇 PDF/链接，产出五段研读卡——观点逐句挂页码锚、复现清单（数据/代码/参数缺口如实标）、引用网络只列检索到的、综述脚手架留你自己的论点槽。适合「这篇论文讲什么」「帮我读这篇 PDF」「综述取材」「引用对不对」等场景。Do NOT use for: 代写或降重改写论文、编造不存在的引文与结论、替你做学术评价。Triggers: 论文研读 / 读这篇 PDF / 文献速览 / 综述骨架 / 引用核对 / 复现清单； paper reading / li | |
| 定增条款 A股定增／再融资条款速读：输入一份定增方案公告（正文、PDF 或披露页定位），产出十项条款逐字卡——发行方式、定价基准日、发行价或价格下限、发行数量与募资上限、发行对象与认购方式、限售期、募集资金用途、摊薄与回报测算、业绩承诺与补偿、审批链条与当前节点——每项挂原句＋锚，缺的那项点名「未检索到」。摊薄影响、参与值不值、审批何时走完一律挂 [待人工]，不预测、不代答。Do NOT use for: 投资建议或参与决策、未检索到的条款编造或补全、收益与风险承诺。Triggers: |
37,101–37,150 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.