能做什么
Live YouTube and Twitch data: channel stats, growth history, monetization, comments and keywords
作者提供的原文
- 连接方式 在云端运行
- 账号或密钥 需要 Authorization
- 分类 营销与社交媒体
可搭配使用
教 AI 做这类工作的 Skill,可与本项一起使用
为什么排在这里
此项在“连接 AI 的应用”标签中,该标签不使用星标,按以下顺序排列
| 1. 在 Claude 连接器目录中 | 否 | |
|---|---|---|
| 2. 团队实测过 | 否 | |
| 3. 适合泰国用户 | 否 | |
| 4. 不计星标的使用量 | 20/100 | 热度 0 · 新鲜度 12 · 仍在响应 8 · 用户喜欢 +0 |
| 5. 登记日期 | 2026-11-18 |
用户喜欢: 0 · 最多加 5 分(每人 1 分),不能用钱买。
质量: 基础检查通过 4/4 项(仅展示,不用于排序)
安装
选择你使用的 AI 应用,然后复制设置。
用一句话让 AI 帮你安装
请帮我安装 Passive Array。按照 https://mcp.popnx.com/i/passivearray-mcp.md 操作,先把命令给我看,等我同意后再运行。粘贴到能在你电脑上工作的 AI(如 Claude Code、Cursor)。它会读取我们的步骤,先给你看命令,等你同意。无需密钥,无需注册。 · 查看 AI 会读取的全文
通过 4/4 项检查 · 你的提问会经过开发者的服务器(passivearray.com) · ✓ 命令已检查 2026年11月18日
第一步:你用哪个 AI?
选择你用的应用,步骤会显示在下方。
可连接 Claude · 约 5 分钟
需要 Passive Array 的密钥(在开发者网站创建) · 在 claude.ai 或桌面应用设置一次,即可在聊天中使用 · 免费账户可添加 1 个 · Team/Enterprise 方案由所有者(Owner)添加
打开 Claude 的 Connectors 页面
CustomizeConnectors
点击“Add” → “Custom” → “Web”,填写这 2 项
把这里的内容复制到 Claude 中同名的输入框,然后点击 Continue
Name · 名称Passive ArrayRemote MCP server URL · 地址https://passivearray.com/mcp在 Request headers 下添加这个名称和值
Header 名称Authorization值Bearer <在这里粘贴你的密钥>只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
在 Claude 检测到的登录设置页点击 Continue · Authentication 按作者说明选择(Sign in now、Sign in when needed 或 No sign in),然后点击 Add
在聊天中开启
+ConnectorsPassive Array
试一试
用 Passive Array 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude 说正在使用 Passive Array,或回答中出现 Passive Array 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
ChatGPT 暂时无法连接 Passive Array
Passive Array 连接时需要你自己的密钥,但 OpenAI 帮助页面只说明了 OAuth 登录,没有我们能核实的填密钥方式
你可以改用这些方式
A · 使用能连接 Passive Array 的应用
- 需要 Passive Array 的密钥(在开发者网站创建) · 在 claude.ai 或桌面应用设置一次,即可在聊天中使用 · 免费账户可添加 1 个 · Team/Enterprise 方案由所有者(Owner)添加
- 需要 Passive Array 的密钥(在开发者网站创建)
- 需要 Passive Array 的密钥(在开发者网站创建)
- 需要 Passive Array 的密钥(在开发者网站创建)
B · 营销与社交媒体 类中能连接 ChatGPT 的其他服务器
从同类中使用较多的里挑选,仅限用地址即可连接、无需密钥的
- Google Ads + Meta Ads + SEO/GEO MCP✓ ChatGPT
- Cargo✓ ChatGPT
- NotFair✓ ChatGPT
可连接 Claude Code · 约 5 分钟
需要 Passive Array 的密钥(在开发者网站创建)
打开终端
Mac:按 Command+空格,输入 Terminal · Windows:打开 PowerShell
粘贴这条命令并按 Enter
claude mcp add --transport http --scope user passivearray https://passivearray.com/mcp --header 'Authorization: Bearer {api_key}'按 Enter 前,把“your-value”改成你的密钥
只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
在 Claude Code 中输入 /mcp
应能看到 Passive Array 在列表中并显示已连接
试一试
用 Passive Array 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Claude Code 说正在使用 Passive Array,或回答中出现 Passive Array 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同
可连接 Cursor · 约 5 分钟
需要 Passive Array 的密钥(在开发者网站创建)
填写你的密钥
在 Passive Array 的设置中,Authorization 还是空的,填入你的密钥并保存
只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
在 Cursor 中以 Agent 模式打开 AI 聊天
试一试
用 Passive Array 告诉我你能帮我做什么,然后做一件给我看✓ 如果 Cursor 说正在使用 Passive Array,或回答中出现 Passive Array 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 Cursor 帮助页面核实 · 你的界面可能因版本略有不同
可连接 VS Code · 约 5 分钟
需要 Passive Array 的密钥(在开发者网站创建)
填写你的密钥
在 Passive Array 的设置中,Authorization 还是空的,填入你的密钥并保存
只把密钥粘贴到应用的输入框里,不要粘贴到聊天中或发给任何人
在 VS Code 中以 Agent 模式打开 AI 聊天
试一试
用 Passive Array 告诉我你能帮我做什么,然后做一件给我看✓ 如果 VS Code 说正在使用 Passive Array,或回答中出现 Passive Array 的工具名称,就表示连接成功
菜单名称已于 2026年11月11日 对照 VS Code 帮助页面核实 · 你的界面可能因版本略有不同
开发者专用:终端命令和 JSON 设置
通过网页连接
Claude Code
claude mcp add --transport http --scope user passivearray https://passivearray.com/mcp --header 'Authorization: Bearer {api_key}'Cursor (~/.cursor/mcp.json)
{
"mcpServers": {
"passivearray": {
"url": "https://passivearray.com/mcp",
"headers": {
"Authorization": "Bearer {api_key}"
}
}
}
}VS Code (mcp.json)
{
"servers": {
"passivearray": {
"type": "http",
"url": "https://passivearray.com/mcp",
"headers": {
"Authorization": "Bearer {api_key}"
}
}
}
}需要你填写的值
- Your free Passive Array API key, from https://passivearray.com/account/#mcp
说明仍为英文原文。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
一览
哪些应用可以使用它
依据这个服务器的连接方式和各应用自己的文档,检查于 2026年11月10日。
通过互联网连接,适用于
ClaudeGooseLM StudioMsty StudioCherry StudioCursorVS Code + GitHub CopilotVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantClaude CodeCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatOpen WebUIRaycast
它能做什么
工具名称和描述与服务器在 2026年11月18日 报告的完全一致。这些内容由作者撰写,不是我们写的;我们以纯文本形式显示。
get_channel_statsLive public stats for a YouTube channel from the official YouTube Data API: subscribers, total views, video count, country, creation date, keywords, topics, and the last 10 uploads with views, likes, comments, length and date, plus averages (views, likes, comments, uploads per mo
get_channel_growthDaily subscriber, view and video-count history for a YouTube channel as recorded by Passive Array, with 1, 7, 30, 90 and 365-day change. History starts the day Passive Array first tracked the channel; nothing earlier is backfilled or estimated. Looking a channel up adds it to dai
check_monetizationEstimates whether a YouTube channel is in the YouTube Partner Program from public signals: subscriber thresholds, uploads in the last 90 days, made-for-kids flags and, when visible, ad placements on recent videos. Returns a verdict with a confidence percentage, never a certainty,
compare_channelsSide-by-side live stats for two or three YouTube channels: subscribers, total views, videos, average views, likes and comments on recent uploads, upload pace and Shorts share. Costs 2 Passive Array credits.
get_video_statsLive stats for up to 50 YouTube videos: title, channel, publish date, views, likes, comments, length, whether it is live, tag count and tags, plus the channel's subscriber count. Costs 1 Passive Array credit.
get_video_commentsPublic comments on a YouTube video or Short, with author, text, likes, reply count and date, plus up to 5 replies per comment. Comments held for review, flagged as spam or deleted are never returned by YouTube's API. Costs 2 Passive Array credits.
analyze_video_commentsComment analytics for one public YouTube video or Short: the channel's reply rate, questions without a reply from the channel (most liked first, with links), buying and collab comments, scam spam still visible, busiest comment hours (UTC) and the words viewers use most. Simple vi
audit_channel_commentsAudits a channel's comment section across its latest 5 or 10 uploads: a 0 to 100 comment health score (Passive Array's rubric, every part returned), reply rate, unanswered questions across videos, regular commenters who comment on several videos, buying and collab comments, visib
keyword_researchMeasured YouTube keyword research: YouTube search autocomplete suggestions for the phrase, the top 10 videos ranking for it today (views, channel size, age, engagement), competition metrics and an opportunity score with its formula. No search-volume number is returned, because no
get_outlier_videosVideos that are getting far more views than their channel's usual upload (views divided by the channel's median), from Passive Array's daily scan of tracked channels. Filter by topic, long-form or Shorts, and age. Good for spotting ideas that are working right now. Costs 1 Passiv
estimate_earningsEstimates a channel's monthly and yearly YouTube ad earnings as a range: views on uploads from the last 30 days times an RPM range (what creators keep per 1,000 views). Excludes sponsorships, memberships and merchandise. Costs 1 Passive Array credit.
search_channelsFinds YouTube channels by keyword or niche, optionally limited to a country (ISO code such as US, GB, IN, BD, PK) and a subscriber range. Returns up to 25 channels with subscribers, views, videos and country. Costs 5 Passive Array credits.
get_twitch_channelA Twitch streamer's live data from the official Twitch Helix API (exact follower total, partner status, live status and viewers, recent videos) plus the daily follower history Passive Array has recorded, with 7, 30 and 90-day change. Costs 1 Passive Array credit.
基础检查i
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月20日
- ✓ 通过
作者与仓库一致
注册名称已通过域名 passivearray.com 验证(注册中心会核实域名所有权)
- ✓ 通过
仍在维护
代码最后更新于昨天
- – 不适用
有许可证
不适用:由 passivearray.com 的所有者自己运行的云端服务器,你不会运行仓库里的代码
- ✓ 通过
端点有响应
在 147 毫秒内响应(HTTP 200)
- – 不适用
使用 OAuth 登录
无需登录
- ✓ 通过
工具描述中没有隐藏指令
已读取 13 个工具描述,未发现针对 AI 的隐藏指令
作者:申请重新检查或报告信息错误
给作者:在 README 中加徽章
如果你是作者,可把这段代码贴到 README。徽章会链接回本页。
[](https://mcp.popnx.com/en/servers/passivearray-mcp)讨论
提问、分享技巧或报告无法使用。按“有帮助”数量排序。
用户评价
还没有评价
与“营销与社交媒体”中的其他服务器比较i
- 星标数多于此分类中其他 4243 个服务器的 0%
- 发布到注册中心的时间比此分类中其他 4243 个服务器的 92% 更新
- 基础检查结果等于或优于此分类中其他 4243 个服务器的 99%
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
常见问题
Passive Array 安全吗?
它在 4 项适用的基础检查中通过了 4 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
它需要哪些密钥或代码?
你需要填写:Authorization
它在我的电脑上运行还是在云端?
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
它还在维护吗?
GitHub 仓库最后更新于昨天,最新版本(1.0.1)于前天进入注册中心。
哪些应用可以使用它?
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,因此也能用于接受远程 MCP 的应用,例如网页版 ChatGPT(Business 及以上为完整 MCP;Pro 为只读)。
有什么可以替代它?
同一分类中有:Relaticle, ChatbotX, X (Twitter) MCP Server by Xquik, Betterlytics, BibiGPT YouTube Video Summarizer
功能相近的其他选择
同一分类、其他作者、命令已通过检查。官方的排在前面,再按星标排序。
如何看卡片:↓ 每周下载量i★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
Relaticle
Open-source CRM your AI agents can write to: companies, people, opportunities, tasks and notes.
ChatbotX
Omnichannel customer messaging, CRM contacts, flows, sequences, broadcasts, and AI automation.
X (Twitter) MCP Server by Xquik
Xquik is the world's fastest and cheapest X (Twitter) data API with the most complete X data.
Betterlytics
Query your Betterlytics web analytics from AI agents: traffic, funnels, journeys, errors, uptime.
BibiGPT YouTube Video Summarizer
AI video summarizer: summarize and transcribe YouTube, Bilibili, TikTok and podcast videos.
| 本项 Passive Array | Relaticle | ChatbotX | X (Twitter) MCP Server by Xquik | Betterlytics | BibiGPT YouTube Video Summarizer | |
|---|---|---|---|---|---|---|
| 运行位置 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 | 在云端运行,无需安装 |
| 需要密钥 | 是 | 否 | 否 | 否 | 是 | 否 |
| 可在 Claude 中按 + 添加 | 否 | 否 | 否 | 否 | 否 | 否 |
| 基础检查 | 4/4 | 4/4 | 4/4 | 4/4 | 4/4 | 4/4 |
| 许可证 | MIT | AGPL-3.0 | NOASSERTION | MIT | AGPL-3.0 | MIT |
| 价格 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 | 登记中未列出费用 |
作者的 README
打开 README
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Passive Array
Free tools for creators and brands. One repository, one Netlify site.
- 35 creator tools at
/creator-tools/: live YouTube and Twitch checks, Instagram, TikTok and X calculators, fake-follower estimators, comparisons and content generators. - Brand kit in
passive-array-brand/: logo, favicons, social images, brand guide.
Deploy on Netlify
- Push this folder to a GitHub repository.
- In Netlify: Add new project, Import an existing project, pick the repository.
- Netlify reads
netlify.toml: build commandnode build.js, publish folderdist. Leave them and click Deploy. - Project configuration, Environment variables, add:
| Variable | Needed for | Where to get it |
|---|---|---|
YOUTUBE_API_KEY | The 9 live YouTube tools | Google Cloud, YouTube Data API v3, free. Steps in creator-tools/README.md |
TWITCH_CLIENT_ID, TWITCH_CLIENT_SECRET | The 2 live Twitch tools | dev.twitch.tv/console, free |
LEMONSQUEEZY_API_KEY, LEMONSQUEEZY_STORE_ID, LEMONSQUEEZY_VARIANT_ID, LEMONSQUEEZY_WEBHOOK_SECRET | The Pro plan | Lemon Squeezy, see the Pro plan section below |
ANTHROPIC_API_KEY | Optional. AI-written hashtags, bios, ideas, growth plan | console.anthropic.com, paid. Without it the generators use built-in templates |
- Trigger a deploy after adding variables. Every later push to
mainredeploys automatically.
Deploy on Vercel instead
The repository also works on Vercel. Import it, keep the detected settings (vercel.json sets the build command and the dist output folder), add the same environment variables under Settings, Environment Variables, and deploy. The API routes run as Vercel Functions from the api/ folder; vercel.json maps the tool URLs onto them, so the pages need no changes.
Research section
/research/ is YouTube keyword research built only from measured sources: YouTube autocomplete (free, no quota) plus the top 10 results for the phrase from the Data API. It shows competing videos, the top 10 with live views and channel sizes, an opportunity score with its formula, and every phrase YouTube autocompletes around the keyword. It never shows a search volume figure, because nobody outside Google has one for YouTube.
Code: lib/research.js (logic), api/research.js and netlify/functions/research.mjs (routes), pages/research-pages.js (page), creator-tools/public/research.js (browser side).
Quota and the key-value store
One keyword analysis costs 102 YouTube quota units; a free key has 10,000 a day. Results are cached for 7 days and a daily guard stops fresh analyses at RESEARCH_DAILY_UNITS (default 8,000), so the other live tools keep working. Without a store the cache lives only in each function instance and the guard resets whenever Vercel starts a new one, so add the store before promoting the page:
- In Vercel open the project, Storage, Create Database, choose Upstash (Redis), free plan, and connect it to the project. Vercel adds
KV_REST_API_URLandKV_REST_API_TOKENautomatically. Alternatively create a database at upstash.com and addUPSTASH_REDIS_REST_URLandUPSTASH_REDIS_REST_TOKENyourself. - Redeploy. Cached keywords then survive across instances, the quota counter is shared, and every analysed keyword starts filling
idx:channels, the channel index the upcoming outlier and growth feeds are built from.
The same store already powers sign-ups and rate limits (lib/store.js).
Outlier videos and Shorts
/research/outliers/ and /research/shorts/ list uploads from the last 90 days that did 3x or more what their channel's other recent uploads did (views / median views of the channel's other last-10 uploads). Filters by topic, minimum multiplier, age and sort, plus a thumbnail-grid view.
The feed is built by a daily job, api/research-cron.js, scheduled in vercel.json (crons, 05:15 UTC, the once-a-day cadence the free plan allows). It walks the channel index (idx:channels), scans up to OUTLIER_CHANNELS_PER_RUN channels (default 250) within OUTLIER_UNITS_PER_RUN quota units (default 3,000), writes the feed to outliers:latest and one snapshot per channel per day (snap: : ) for the growth rankings. On an empty index it first seeds a dozen broad niches. A store lock stops it running more than once every 6 hours.
Optional: set CRON_SECRET so only Vercel's scheduler (and you, with /api/research-cron?key=...) can trigger the job. Logic: lib/outliers.js.
How it is built
build.js Builds dist/: home page, creator tools, blog, company pages, brand files
netlify.toml Netlify settings and function packaging netlify/functions/ API routes on Netlify (YouTube, Twitch, AI) api/ The same routes for Vercel; vercel.json rewrites the tool URLs to them creator-tools/ tools.js (the list of tools), build-tools.js (page generator), public/ (CSS and JS), lib/ (API helpers) passive-array-brand/ Logo SVGs, favicons, social images, brand-guide.html, make-brand.js
Run locally without Netlify:
node build.js # builds dist/ so you can open dist/index.html
creator-tools/Start.bat # local server for the creator tools with live API routes
Adding a creator tool
- Add an entry to
creator-tools/tools.js. - Add
COMPUTE["your-slug"]increator-tools/public/shared.js. - Run
node build.js.
Details in creator-tools/README.md. The brand guide is passive-array-brand/brand-guide.html.
Notes
- Never commit API keys. Local keys go in
creator-tools/config.json, which git ignores. - The video downloader from the original workspace is not included: it needs ffmpeg and yt-dlp binaries that Netlify cannot run.
Site structure (beyond the tools)
| Path | Source | Notes |
|---|---|---|
/blog/ and /blog/ / | blog/posts/*.md, built by blog/build-blog.js | Add an article by adding a Markdown file with the front matter shown at the top of build-blog.js. Newest first. |
/about/, /contact/, /privacy/, /terms/ | pages/site-pages.js | Company pages. |
/youtube-extension/ and /youtube-extension/privacy/ | youtube-extension/pages.js | Landing page and privacy policy for the Chrome extension. |
chrome-extension/ | The extension source itself | See chrome-extension/README.md and STORE-LISTING.md. Not part of the website build. |
/sitemap.xml, /robots.txt | generated by build.js | Every tool, article and page. |
/api/subscribe | api/subscribe.js | Receives sign-ups, newsletter and contact forms. |
Sign-up, newsletter and contact forms
The "Sign up free" button, the footer newsletter box and the contact page all POST to /api/subscribe, which forwards each entry as JSON to a destination you own. Nothing is stored on the site itself.
Until a destination is set, the forms say "Sign-up is not open yet." The buttons and the modal work; there is simply nowhere to put the email. Check the current state at any time:
https://passivearray.vercel.app/api/subscribe?health=1
It answers "configured": true or false. It never reveals the destination.
Option A: Google Sheet (recommended, about five minutes)
Always reachable, free, and needs no account beyond the Google one you have.
- Create a new Google Sheet. Name it something like "Passive Array sign-ups".
- In that sheet choose Extensions, Apps Script. A code editor opens.
- Delete whatever is in
Code.gs, then paste the whole contents ofsetup/google-sheet-receiver.gsfrom this repository. Click the save icon. - Click Deploy, New deployment. Press the gear next to "Select type" and choose Web app.
- Set Execute as to Me, and Who has access to Anyone. This is required: Vercel calls it as an anonymous visitor. The URL is unguessable and the script only ever appends rows.
- Click Deploy. Google asks you to authorise the script the first time. Approve it. On the "Google hasn't verified this app" screen choose Advanced, then Go to (your project name). It is your own script.
- Copy the Web app URL. It looks like
https://script.google.com/macros/s/AKfy.../exec. - Paste that URL into a browser. It should answer
{"ok":true,"service":"Passive Array sign-up receiver"}. If it asks you to log in, "Who has access" is not set to Anyone. - In Vercel open the project, Settings, Environment Variables. Add
SUBSCRIBE_WEBHOOK_URLwith that URL. Apply it to Production. - Go to Deployments and Redeploy the latest one. Environment variables only reach a deployment when it is built.
- Open
/api/subscribe?health=1. It should now say"configured": true. - Sign up on the site with your own email. A row appears in the sheet within a second or two, with the header row created automatically.
Contact-form messages also email you, so you see them the same day. Delete the MailApp.sendEmail block in the script if you would rather they only appear in the sheet.
Option B: n8n
Use this if your n8n is reachable from the internet. A self-hosted n8n on your own PC is not: Vercel cannot call localhost. n8n Cloud, or a self-hosted instance behind a public domain or a tunnel, works fine.
- New workflow, add a Webhook node. Method POST, path
passive-array. - Copy the Production URL, not the Test URL. The test URL only listens while you have the editor open.
- Add a Google Sheets node (append row) or whatever storage you prefer, mapping the fields below.
- Add a Respond to Webhook node returning any 200 response.
- Activate the workflow. An inactive workflow rejects production calls.
- Set
SUBSCRIBE_WEBHOOK_URLin Vercel to the Production URL and redeploy.
What gets sent
Any service that accepts a JSON POST works. The body is:
{ "kind": "signup", "email": "someone@example.com", "name": "", "platform": "YouTube",
"message": "", "source": "/blog/", "receivedAt": "2026-09-22T10:00:00.000Z" }
kind is signup from the button, newsletter from the footer box, or contact from the contact page. Emails arrive lowercased and trimmed. A hidden honeypot field blocks the common bots before anything is forwarded.
If a sign-up fails
The visitor sees "Could not save that right now." The reason is written to the Vercel function logs: open the project, Logs, and filter for subscribe. The usual causes are an inactive n8n workflow, an Apps Script deployment whose access is not set to Anyone, or a URL that was saved without redeploying.
Accounts and sign-in
Sign-in is passwordless. There are no passwords stored and no user database: a link is signed with a secret, emailed, and exchanged for a session cookie. The account page holds the person's email address and one preference, nothing more, and every tool on the site works without an account.
Turning it on
- Set up the sign-up destination first (the section above). The same Apps Script delivers the sign-in emails, so no extra service is needed.
- Generate a secret:
node -e "console.log(require('crypto').randomBytes(32).toString('hex'))"
- In Vercel add the environment variable
AUTH_SECRETwith that value, andSITE_URLset tohttps://passivearray.vercel.appso the links point at the right place. Then Redeploy. - Check it:
https://passivearray.vercel.app/api/auth?action=healthshould answer"enabled": true, "canEmail": true. - Go to
/login/, enter your own email, and click the link that arrives.
Changing AUTH_SECRET later signs everybody out, which is also how you would respond if you ever thought it had leaked.
How it works
| Step | What happens |
|---|---|
| Ask for a link | /api/auth?action=request signs {email, expiry} with AUTH_SECRET and sends it to the webhook as kind: "login" |
| Email | The Apps Scri