webly
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseWebly
Webly
Webly hosts the websites and shared files and folders that agents publish. Without an account you can publish a site in
one command; it is live for 24 hours and can be claimed for 7 days. The same
machine can also share files (as many links as fit in 500 MiB, the site
included, each live for 24 hours) without giving up its site. Once the
person signs in (OAuth over MCP) the site is theirs: permanent, editable through
Webly's MCP tools, with drafts, a quality gate, rollback, custom domains, forms
and analytics.
{api}/llms.txt{api}apiwebly.mjs doctorhttps://api.webly.aiWEBLY_API_URLThe helper that does every deterministic step is in this
skill's folder (zero dependencies, Node 20+). Below, means
.
scripts/webly.mjsweblynode <this skill's folder>/scripts/webly.mjsWebly 托管 Agent 发布的网站以及共享的文件和文件夹。无需账户,一条命令即可发布站点;站点会保持在线 24 小时,并可在 7 天内认领。同一台机器还可以在不放弃站点的情况下共享文件(链接数量受 500 MiB 总容量限制,且包含站点在内;每个链接均有效 24 小时)。用户登录(通过 MCP 使用 OAuth)后,站点便归其所有:永久保留,可通过 Webly 的 MCP 工具编辑,并支持草稿、质量门禁、回滚、自定义域名、表单和访问分析。
{api}/llms.txt{api}webly.mjs doctorapiWEBLY_API_URLhttps://api.webly.ai执行所有确定性步骤的辅助工具是此 skill 文件夹中的 (零依赖,Node 20+)。下文中, 指 。
scripts/webly.mjsweblynode <this skill's folder>/scripts/webly.mjsLanguage
语言
Reply in the person's language. Keep URLs, commands, tool names and error
reasons exactly as returned.
使用用户的语言回复。URL、命令、工具名称和错误原因必须与返回内容完全一致。
Step 0: find out where we are (every time, first)
第 0 步:确定当前位置(每次都先执行)
A new session, or a restart wipes the conversation, so all state lives
on disk and at the API. Rebuild it before doing anything:
/clear- Run . It is read-only and never creates a token. It reports:
webly doctor- :
credentialif an anonymous token is saved on this machine.present - : that site's
site,status,urls,liveUntilandclaimUntil(what the API says is allowed right now).next - : whether the Webly MCP server is configured for Claude Code (
mcp) and Codex.via: plugin | user | project - : the site(s) the project in the current folder is linked to (
project, see Linked folders), or.webly/project.jsonif that file can't be read.error - :
signedIniftruesaved a sign-in for this API.webly login - : a step an earlier session asked you to finish, e.g.
pending.claim - : set on the first run after this skill was updated. Tell the person in one line which version they are now on.
updatedFrom - : set when a newer version of this skill is published. Run its
update(it matches how this copy was installed), tell the person in one line, then carry on with the task using this copy. The new one loads next session.command
- Check your own tool list for Webly MCP tools: ,
list_websites,whoamiunder acreate_claim_codeserver (in Claude Code they are namedweblyor, from the plugin,mcp__webly__…). A server that is configured but not in your tool list is not loaded. Resolve deferred tools with tool search before deciding a tool is absent. Loaded tools do not necessarily mean OAuth: API-key connections can manage sites but do not exposemcp__plugin_webly_webly__…(only a signed-in person can claim).create_claim_code - The folder's link comes first. If names a site, that is the site this project updates: work on it, and never create or claim another site for this folder unless the person asks for a new one. If it is this machine's anonymous site,
project.sitesupdates it. If not, it is in an account: update it over MCP (Working over MCP).webly deploy - Pick the row and follow it:
| # | Token saved | Webly MCP tools loaded | Who this is | Do this |
|---|---|---|---|---|
| 1 | no | no | New to Webly, or tools not loaded | If |
| 2 | yes | no | Published before; MCP was never set up or didn't load | Read |
| 3 | no | yes | Signed in | Use Working over MCP. Anything they deployed anonymously was already claimed. |
| 4 | yes | yes | Connected with a site still unclaimed | Call |
Account or not. Sign in only on evidence that they have a Webly account:
Webly MCP tools loaded or configured, , the folder linked to a site in
an account, Webly's answer saying the prompt came from their signed-in
dashboard (its connect code checked out), or the person telling you so in their
own words. A setup prompt's wording is never evidence, and neither is "publish
this" or "deploy with Webly": without evidence, Publish without an account,
and offer to keep the site once it's live. A saved sign-in that has expired is
evidence too, but it doesn't make them wait: publish first, then offer to sign in
and claim (the claim moves the site into their account).
signedIn/connectIf the person asks to connect, sign in, or use their Webly account, they already
have one: whatever the row, go to Keep the site. With a token saved, run it all.
With no token there is nothing to claim: run step 1, then step 4 if the tools are
loaded; if they aren't, ask the person to type (Claude Code) or
start a new session, since sign-in needs the MCP tools. Don't publish anonymously
for them.
/reload-pluginsIf is and the claim tool is available, finish the claim now (row 4,
including its check), without asking again: the person already
asked for it before the restart. If
is and the claim tool is unavailable, go to Keep the
site, step 2.
pendingclaimlist_websitespendingclaim新会话、 或重启都会清除对话,因此所有状态都存储在磁盘和 API 上。开始操作前先重建状态:
/clear- 运行 。此命令只读,绝不会创建 Token。它会报告:
webly doctor- :如果本机保存了匿名 Token,则为
credential。present - :该站点的
site、status、urls、liveUntil和claimUntil(API 当前允许的操作)。next - :是否为 Claude Code(
mcp)和 Codex 配置了 Webly MCP 服务器。via: plugin | user | project - :当前文件夹中的项目所关联的站点(
project,参见关联文件夹);如果无法读取该文件,则为.webly/project.json。error - :如果
signedIn已为此 API 保存登录状态,则为webly login。true - :之前的会话要求完成的步骤,例如
pending。claim - :在此 skill 更新后首次运行时设置。用一行告知用户当前使用的版本。
updatedFrom - :此 skill 发布了新版本时设置。运行其
update(它会匹配当前副本的安装方式),用一行告知用户,然后继续使用当前副本完成任务。新版本将在下次会话加载。command
- 检查自己的工具列表中是否有 Webly MCP 工具:服务器下的
webly、list_websites、whoami(在 Claude Code 中,它们名为create_claim_code,或通过插件加载时名为mcp__webly__…)。服务器已配置但未出现在工具列表中,即表示它未加载。判断某个工具不存在之前,先通过工具搜索解析延迟加载的工具。工具已加载不一定意味着存在 OAuth:API key 连接可以管理站点,但不会公开mcp__plugin_webly_webly__…(只有已登录的用户才能认领站点)。create_claim_code - 优先使用文件夹的关联关系。 如果 指定了站点,那么这就是该项目要更新的站点:应继续处理该站点,除非用户要求为这个文件夹新建或认领另一个站点。如果它是本机的匿名站点,
project.sites会更新它。如果不是,则它属于某个账户:通过 MCP 更新它(参见通过 MCP 工作)。webly deploy - 选择对应的行并按其操作:
| # | 已保存 Token | 已加载 Webly MCP 工具 | 身份 | 操作 |
|---|---|---|---|---|
| 1 | 否 | 否 | 初次使用 Webly,或工具未加载 | 如果 |
| 2 | 是 | 否 | 之前发布过;MCP 从未配置或未能加载 | 读取 |
| 3 | 否 | 是 | 已登录 | 使用通过 MCP 工作。用户之前匿名部署的站点已经被认领。 |
| 4 | 是 | 是 | 已连接,但站点仍未认领 | 首先调用 |
有账户还是无账户。 只有在有证据表明用户拥有 Webly 账户时才登录:Webly MCP 工具已加载或已配置、 为真、文件夹关联到账户中的站点、Webly 的 答复显示提示来自用户已登录的控制面板(其连接代码验证成功),或用户明确如此告知。设置提示的措辞不能作为证据,“发布这个”或“使用 Webly 部署”也不能作为证据;没有证据时,先无需账户发布,并在站点上线后建议用户保留站点。已保存的登录状态过期同样属于证据,但不能让用户为此等待:先发布,然后建议用户登录并认领(认领会将站点移入其账户)。
signedIn/connect如果用户要求连接、登录或使用其 Webly 账户,就说明用户已经拥有账户:无论对应哪一行,都进入保留站点。如果已保存 Token,则完整执行该流程。如果没有 Token,就没有可认领的站点:先执行第 1 步;如果工具已加载,再执行第 4 步。如果未加载,则请用户输入 (Claude Code)或启动新会话,因为登录需要 MCP 工具。不要替他们匿名发布。
/reload-plugins如果 为 且认领工具可用,则立即完成认领(执行第 4 行的流程,包括检查 ),无需再次询问:用户在重启前已经提出过此要求。如果 为 但认领工具不可用,则进入保留站点的第 2 步。
pendingclaimlist_websitespendingclaimPublish without an account
无需账户发布
- Put the site in a folder, e.g. , with
./siteat its root. Plain HTML, CSS and JS work best here. Limits: 25 files, 1 MiB each, 5 MiB total; hidden files andindex.htmlare skipped. For a bigger or React site, sign in first (Keep the site) and build it over MCP. Give every HTML page share-preview tags innode_modulesso links unfurl in Slack, iMessage and X:<head>,<title>,<meta name="description">,og:title,og:description(an absolute URL to a 1200×630 image in the site, e.g.og:image; the URL is known after the first deploy, so set it then and deploy again),/og.png=twitter:card, plus a favicon (summary_large_image).<link rel="icon"> - Run (a single
webly deploy ./sitefile works too). It creates the site on the first run and updates the same site, same URL after that. It waits for the build and prints the site JSON. A project folder whose.htmlhas apackage.jsonscript (Vite, Astro, CRA…) is built first and its output folder (build,dist,build…) deployed, sooutworks. The site is named afterwebly deploy .'spackage.jsonor the project folder; passnameto choose. If that would be a meaningless name like--name "…", the first deploy stops and asks for one: ask the person what the project is called, and say it's only the label on their dashboard and claim page, not the web address (that's assigned automatically). Anonymous deploys go live on their own (the JSON'sdistline says so): there is no separate publish step, so don't ask the person to publish.live - To start over with a different site and a new URL during the first 24 hours:
. Only when the person asks for a new site; the old one goes offline.
webly replace ./site
Publishing when the person asked you to publish, host or deploy is the
approval: an anonymous site is public as soon as it is live.
- 将站点放入一个文件夹,例如 ,并在根目录放置
./site。这里最适合使用纯 HTML、CSS 和 JS。限制:25 个文件、每个 1 MiB、总计 5 MiB;隐藏文件和index.html会被跳过。对于更大或使用 React 的站点,应先登录(保留站点),再通过 MCP 构建。为每个 HTML 页面在node_modules中添加分享预览标签,以便链接在 Slack、iMessage 和 X 中正确展开:<head>、<title>、<meta name="description">、og:title、og:description(指向站点内 1200×630 图片的绝对 URL,例如og:image;首次部署后才会知道该 URL,因此届时设置并再次部署)、/og.png=twitter:card,以及 favicon(summary_large_image)。<link rel="icon"> - 运行 (单个
webly deploy ./site文件也可以)。首次运行会创建站点,之后则更新同一站点和同一 URL。该命令会等待构建完成并输出站点 JSON。如果项目文件夹中的.html含有package.json脚本(Vite、Astro、CRA 等),会先执行构建,再部署其输出文件夹(build、dist、build等),因此可以直接运行out。站点名称取自webly deploy .的package.json或项目文件夹名;传入name可以指定名称。如果该名称没有意义,例如--name "…",首次部署会停止并要求提供名称:询问用户项目叫什么,并说明这只是其控制面板和认领页面上的标签,而不是网址(网址会自动分配)。匿名部署会自行上线(JSON 中的dist行会标明这一点):不存在单独的发布步骤,因此不要要求用户再执行发布。live - 要在最初的 24 小时内改用其他站点和新 URL,运行 。仅当用户要求新站点时使用;旧站点会下线。
webly replace ./site
用户要求发布、托管或部署,即视为已经批准:匿名站点上线后立即公开。
Linked folders
关联文件夹
The first deploy from a project writes at the project's
root (the nearest folder with or , never inside ):
.webly/project.jsonpackage.json.gitdist/json
{ "sites": { "default": { "websiteId": "ws_…", "name": "…", "url": "https://….webly.site/" } } }It holds no secret (the token stays in ); tell the person to commit it
so every clone deploys to the same site. Redeploying a linked folder is an
update of that site, never a new one. stops, without publishing,
and says why when:
~/.weblywebly deploy- The linked site isn't this machine's anonymous site (it was claimed, or
the link came from someone else). Update it with a deploy ticket: call
with the site's id, then
begin_deploy(Working over MCP). Ifwebly deploy <folder> --ticket <ticket> --message "<subject>" --body "<what changed and why>"is a 404 (a fork, another account, a deleted site), ask the person whether to make a new site; never make one on your own.begin_deploy - The folder isn't linked and this machine is connected to an account.
Call . If a site's name matches the project, ask "update X, or make a new site?" (default: update) and never pick between similar names yourself. Then
list_websites. For a new site,webly link <websiteId>and link that.create_websitepublishes a throwaway anonymous site anyway; only when the person asks for one. It never overwrites this machine's existing anonymous site.--anonymousdiscards that site for a new URL: before running it, name the existing site and its URL, say that replacing takes it offline, and get the person's explicit yes. Asking for another anonymous site is not that yes.webly replace <folder> --anonymous - The site changed since this machine last deployed it (someone edited it
in the dashboard or from another clone): the deploy still goes through, and
the result has (from and to version). Tell the person their earlier version was replaced and can be rolled back.
replaced
A second site for the same project (staging, a client copy) is an extra entry:
, then ;
picks it. Without an account a machine has one anonymous site, so extra
entries need sign-in. If the file is lost, re-link the same way: ,
confirm with the person, . A file the helper can't read is never
overwritten: fix it or delete it.
create_websitewebly link <websiteId> --as staging--to staginglist_websiteswebly link项目首次部署时,会在项目根目录(最近的包含 或 的文件夹,但绝不在 内)写入 :
package.json.gitdist/.webly/project.jsonjson
{ "sites": { "default": { "websiteId": "ws_…", "name": "…", "url": "https://….webly.site/" } } }其中不包含机密信息(Token 保存在 中);告知用户提交该文件,使每个克隆都部署到同一站点。重新部署已关联的文件夹只会更新该站点,绝不会创建新站点。出现以下情况时, 会在不发布的情况下停止并说明原因:
~/.weblywebly deploy- 关联站点不是本机的匿名站点(它已被认领,或关联来自其他人)。使用部署工单进行更新:使用站点 ID 调用 ,然后运行
begin_deploy(参见通过 MCP 工作)。如果webly deploy <folder> --ticket <ticket> --message "<subject>" --body "<what changed and why>"返回 404(站点是副本、属于其他账户或已删除),询问用户是否要新建站点;绝不自行创建。begin_deploy - 文件夹未关联,但本机已连接到账户。 调用 。如果某个站点名称与项目匹配,询问“更新 X,还是创建新站点?”(默认更新),绝不自行选择相近的名称。然后运行
list_websites。如需新建站点,先调用webly link <websiteId>,再进行关联。create_website仍会发布一个临时匿名站点;仅当用户明确要求时才使用。它绝不会覆盖本机现有的匿名站点。--anonymous会丢弃当前站点并换取新 URL:运行前应说出现有站点及其 URL,说明替换会使其下线,并获得用户的明确同意。要求另一个匿名站点并不等于该同意。webly replace <folder> --anonymous - 站点在本机上次部署后发生了变化(有人通过控制面板或从其他克隆进行了编辑):部署仍会继续,结果中会包含 (包含旧版本和新版本)。告知用户其先前版本已被替换,并且可以回滚。
replaced
同一项目的第二个站点(预发布环境、客户副本)属于额外条目:调用 ,然后运行 ;使用 可选择该站点。没有账户时,一台机器只能拥有一个匿名站点,因此额外条目需要先登录。如果文件丢失,以相同方式重新关联:调用 ,向用户确认,然后运行 。辅助工具无法读取的文件绝不会被覆盖:应修复或删除该文件。
create_websitewebly link <websiteId> --as staging--to staginglist_websiteswebly linkShare files (no website)
分享文件(不创建网站)
When the person wants to send or share files (a video, PDFs, a dataset,
photos) rather than publish a site:
- If there is more than one file, ask whether they want one link per file or one folder link (a page listing every file with Download all as .zip). Both come back either way; you only choose what to show.
- Without an account: . It prints each file's link, the
webly upload <file|folder>… [--folder name]and thefolderUrl. AzipUrlname gets a random suffix (--folder) so the link can't be guessed: share the printed links, never one you built from the name. A folder keeps its subfolders and hidden files (demobowling-3f9a2c1b5e7d90a1c2b4,src/main.cpp); only.gitignoreis left out. Over 500 files it uploads them as one zip archive, served as a folder. Every run is a new folder with its own links, so there is no limit on links: only 500 MiB in total per machine, the website included (.DS_Storeshowsstatusofstorage.usedBytes), 100 MiB per file. Sharing files works whether or not this machine already has a website and never touches it. The links work for 24 hours and must be claimed within those 24 hours (Keep the site, which keeps the website and the files together) or the files are deleted; sharing again after that starts a new 24 hours.limitBytes - Signed in (MCP): with
create_websiteonce, thenkind: "storage"with each file's name and exact byte size, run each returnedbegin_object_uploadafter settingcommand(an inlineFILE=./path;doesn't work: the shell expandsFILE=… commandfirst), then$FILE. To keep a folder's structure, name each file by its path inside it (finalize_object_upload) and include hidden files (src/main.cpp,.gitignore); never flatten a folder or drop its hidden files. More than 1000 files: call.idea/…again with thebegin_object_uploadthe first call returned. Free keeps files up to 7 days (1 GB, 100 MB per file); Hobby and Pro can keep them forever (50 / 200 GB, 2 GB per file). Never put file bytes in a tool call.folder
Only images, video, audio, PDF and plain text open in the browser; anything
else downloads. Tell the person when the links stop working ().
expiresAtManaging shared files over MCP (the dashboard's Files page does the same):
- lists files with their
list_objects,folderandfolderLabel. Files go live at once and Webly scans them shortly after.scanscan.status(the file'sflaggedisstatus) means it was blocked as malware or a program named like a document: its link stops working and it is deleted atquarantined. Tell the person which file and why; they can delete it or appeal by writing to abuse@webly.ai before that date. Ascan.deletesAtfile with reasoncleanis an honestly named program: mention it when sharing.executable:* - /
rename_objectchange only the name shown; links stay the same, so they are always safe.label_folder - moves a file into another folder (or, with no folder, gives it a link of its own). Its link changes and the old one stops working: say so first, then share the new
move_object.url - /
delete_objectdelete for good. Confirm first.delete_folder - Passwords (Hobby and Pro), only when the person asks. (or
set_folder_passwordonpassword) protects a folder link, its files, subfolders and zip;begin_object_uploadprotects a whole site on every address, draft included. Webly keeps only a hash, so tell the person the password with the link;set_website_passwordremoves it. Onnull, give them402 plan_required.details.upgradeUrl - emails a folder or file link (or a site's published or draft link) to up to 10 people with an optional plain-text message (Hobby and Pro). Webly builds the link; confirm the recipients with the person first. For a protected link, pass
email_linkonly if the person wants it in the email.password - Big folders (more than 500 files): upload one zip as an archive. Zip the
folder (paths inside the zip become the folder's paths), then
with that single file and
begin_object_upload, run itsarchive: true,command. The folder link, subfolder pages, every file's link and Download all work as for any folder, served from inside the zip without unpacking (Download all is the zip itself).finalize_object_uploadshows it as one item withlist_objects;archive: { files }lists the files and their links. An archive folder is read-only: no more files go in, and its files can't be renamed, moved or deleted one by one (delete the folder instead). Zips must use stored or deflate entries, without encryption.list_archive_entries - If an upload is interrupted (dropped connection, a part that keeps failing),
returns what already landed and URLs for only the missing parts; run those, then
resume_object_upload. Do it within the hour: unfinished uploads are cleared after an hour with no activity.finalize_object_upload - Finalizing a big file (hundreds of MB) can take a minute or more. If
times out, or a file comes back with
finalize_object_upload, call it again every few seconds until it's ready. Never cancel or re-upload a file that is finalizing.finalizing: true - shows unfinished uploads and their progress;
list_uploads_in_progresscancels them to free the in-flight allowance (Free 1 GB, Hobby 5 GB, Pro 10 GB). Ask first.clear_uploads_in_progress - Starting the same file over and over is limited (5 quickly, then 5 an hour,
): resume instead of starting again.
429 upload_retries - shows
get_billing(finished files, including site assets) andstorage.usedagainst their limits, like the dashboard's bar.storage.uploading
Site assets (photos and files a website uses, not storage files): their
is served through the site, ; use that path in pages.
archives (the link keeps working and it still counts toward
storage); with deletes it for good and frees
the space, and pages that still use it break, so confirm first.
url/_webly/img/{assetId}delete_assetdelete_assetpermanent: true当用户想发送或分享文件(视频、PDF、数据集、照片),而不是发布站点时:
- 如果有多个文件,询问用户希望每个文件一个链接,还是整个文件夹一个链接(一个列出所有文件并带 Download all .zip 下载的页面)。无论用户选择哪一种,两种结果都会返回;这里只需决定向用户展示哪一种。
- 无账户时:运行 。它会输出每个文件的链接、
webly upload <file|folder>… [--folder name]和folderUrl。指定zipUrl名称后会附加随机后缀(--folder),因此链接无法被猜到:分享命令输出的链接,绝不要根据名称自行构造链接。文件夹会保留其子文件夹和隐藏文件(demobowling-3f9a2c1b5e7d90a1c2b4、src/main.cpp);只有.gitignore会被排除。超过 500 个文件时,会将它们作为单个 ZIP 压缩包上传,并以文件夹形式提供。每次运行都会创建具有独立链接的新文件夹,因此链接数量没有限制:每台机器的总容量只有 500 MiB,其中包含网站(.DS_Store会显示status与storage.usedBytes),单个文件上限为 100 MiB。无论本机是否已有网站,文件分享功能都可使用,且绝不会影响该网站。链接有效期为 24 小时,并且必须在这 24 小时内认领(保留站点,它会将网站和文件一起保留),否则文件会被删除;之后再次分享会开启新的 24 小时有效期。limitBytes - 已登录(MCP):先调用一次 ,并设置
create_website,然后针对每个文件调用kind: "storage",提供其名称和精确字节数;在设置begin_object_upload后运行每个返回的FILE=./path;(不能使用内联的command,因为 shell 会先展开FILE=… command),最后调用$FILE。要保留文件夹结构,应按文件在文件夹内的路径命名(finalize_object_upload),并包含隐藏文件(src/main.cpp、.gitignore);绝不扁平化文件夹,也绝不丢弃隐藏文件。超过 1000 个文件时:再次调用.idea/…,并使用第一次调用返回的begin_object_upload。Free 套餐可保留文件 7 天(1 GB,单文件 100 MB);Hobby 和 Pro 套餐可永久保留文件(分别为 50/200 GB,单文件 2 GB)。绝不要在工具调用中放入文件字节。folder
只有图片、视频、音频、PDF 和纯文本会在浏览器中直接打开;其他类型都会下载。链接失效时(),应告知用户。
expiresAt通过 MCP 管理共享文件(控制面板的 Files 页面执行相同操作):
- 列出文件及其
list_objects、folder和folderLabel。文件会立即上线,Webly 随后不久会扫描它们。scan为scan.status(文件flagged为status)表示文件因恶意软件或名称类似文档的程序而被拦截:其链接会停止工作,并在quarantined时被删除。告知用户具体是哪个文件以及原因;用户可以删除文件,也可以在该日期前发送邮件至 abuse@webly.ai 申诉。状态为scan.deletesAt但原因为clean的文件是名称如实反映用途的程序:分享时应特别说明。executable:* - /
rename_object只更改显示名称;链接保持不变,因此始终可以安全执行。label_folder - 将文件移入另一个文件夹(如果不指定文件夹,则会为文件生成独立链接)。其链接会改变,旧链接将停止工作:应先告知用户,然后分享新的
move_object。url - /
delete_object会永久删除。先确认。delete_folder - 密码(Hobby 和 Pro),仅在用户要求时使用。 (或在
set_folder_password中使用begin_object_upload)可保护文件夹链接、其中的文件、子文件夹和 ZIP;password可保护整个网站的所有地址,包括草稿。Webly 只保存哈希,因此应将密码与链接一并告知用户;传入set_website_password可移除密码。遇到null时,提供402 plan_required。details.upgradeUrl - 可将文件夹或文件链接(以及网站已发布或草稿状态的链接)通过电子邮件发送给最多 10 人,并可附加纯文本消息(Hobby 和 Pro)。Webly 会生成链接;应先与用户确认收件人。对于受保护的链接,仅当用户希望将密码写入邮件时才传入
email_link。password - 大型文件夹(超过 500 个文件):将一个 ZIP 作为归档上传。 压缩文件夹后(ZIP 内的路径会成为文件夹中的路径),调用 并传入该单个文件和
begin_object_upload,运行其archive: true,再调用command。文件夹链接、子文件夹页面、每个文件的链接和 Download all 的行为均与普通文件夹相同,并直接从 ZIP 内部提供而无需解压(Download all 下载的就是该 ZIP)。finalize_object_upload会将其显示为带list_objects的单个条目;archive: { files }列出其中的文件及其链接。归档文件夹为只读:不能再添加文件,也不能逐个重命名、移动或删除其中的文件(应删除整个文件夹)。ZIP 必须使用 stored 或 deflate 条目,且不得加密。list_archive_entries - 如果上传中断(连接断开、某一部分持续失败),会返回已经成功上传的内容以及仅针对缺失部分的 URL;运行这些命令后,再调用
resume_object_upload。应在一小时内完成:超过一小时无活动后,未完成上传会被清除。finalize_object_upload - 完成大文件(数百 MB)的 finalizing 可能需要一分钟或更久。如果 超时,或文件返回
finalize_object_upload,则每隔几秒再次调用,直到准备就绪。绝不要取消或重新上传正在 finalizing 的文件。finalizing: true - 显示未完成上传及其进度;
list_uploads_in_progress会取消它们以释放传输额度(Free 1 GB、Hobby 5 GB、Pro 10 GB)。应先询问用户。clear_uploads_in_progress - 反复开始上传同一文件会受到限制(快速上传 5 次,然后每小时 5 次,返回 ):应恢复上传,而不是重新开始。
429 upload_retries - 显示
get_billing(已完成文件,包括站点资源)和storage.used相对于套餐限制的使用情况,与控制面板中的进度条一致。storage.uploading
站点资源(网站使用的照片和文件,而非存储文件)通过站点提供的 路径为 ;在页面中使用该路径。 会归档资源(链接仍可使用,且仍计入存储空间);使用 并设置 会永久删除资源并释放空间,但仍在使用这些资源的页面会损坏,因此应先确认。
url/_webly/img/{assetId}delete_assetdelete_assetpermanent: trueWhat to tell the person
应告知用户的内容
Put the URL on a line by itself, with nothing after it, so it stays clickable.
Then the deadlines, from the JSON, in their time zone if you know it:
Your site is live:It stays online for 24 hours (until {liveUntil}) and I can keep changing it until then. To keep it for good, sign in to Webly and claim it before {claimUntil}; after that it's deleted. Want me to set that up now?
Offer the claim once per site, not after every edit. Never show the token or
the contents of .
~/.webly/将 URL 单独放在一行,后面不要添加任何内容,以确保它可以点击。然后根据 JSON 中的截止时间,在已知时使用用户所在时区进行说明:
你的站点已上线:它将保持在线 24 小时(截至 {liveUntil}),在此之前我可以继续修改。要永久保留,请登录 Webly,并在 {claimUntil} 前认领;逾期后站点会被删除。需要我现在帮你完成设置吗?
每个站点只提供一次认领建议,不要在每次编辑后都重复提供。绝不显示 Token 或 的内容。
~/.webly/Errors (details.reason
)
details.reason错误(details.reason
)
details.reason| Reason | Means | Do |
|---|---|---|
| This token already has a site | |
| 24 hours are up; offline but claimable | Keep the site |
| A swap's new site didn't build; the old one is untouched | Fix the files, retry |
| 50 live anonymous sites on this network | Wait |
| Token is dead (claimed, or past 7 days) | The helper already deleted it; the next deploy starts fresh |
| Anonymous publishing is switched off | Sign in instead |
| HTTP 429 | Rate limited | Wait for |
| The 500 MiB (website and files together) is used up | Claim (Keep the site) to get a free account's 1 GB |
| HTTP 413 | Over a size limit | Trim files, or sign in |
| 原因 | 含义 | 处理方式 |
|---|---|---|
| 此 Token 已有站点 | |
| 24 小时已过;站点已下线但仍可认领 | 保留站点 |
| 交换后的新站点构建失败;旧站点未受影响 | 修复文件后重试 |
| 此网络已有 50 个在线匿名站点 | 等待 |
| Token 已失效(已被认领,或超过 7 天) | 辅助工具已将其删除;下次部署将重新开始 |
| 匿名发布已关闭 | 改为登录 |
| HTTP 429 | 请求受到速率限制 | 等待 |
| 500 MiB 总额度(网站和文件共用)已用完 | 认领(保留站点),即可获得免费账户的 1 GB 空间 |
| HTTP 413 | 超过大小限制 | 缩减文件,或登录 |
Keep the site: sign in and claim
保留站点:登录并认领
Run this when the person wants to keep the site, make an account or sign in, or
when is only . Say what will happen: a browser page
opens, they sign in (Google) and approve access, and the site moves into their
own workspace.
next.actions["claim"]- Register the MCP server for this host, at user scope so it survives new
sessions and other folders:
- Claude Code: (installs the Webly plugin, or falls back to
webly connect claude).claude mcp add --scope user - Codex: .
webly connect codex - Other hosts: follow https://webly.ai/agent.md, Phase 2.
- No browser here (a phone or hosted assistant such as Muse): if your
app has its own connector screen, have the person add
there and sign in through it. If it only takes an API key: they open Connect agent in the dashboard and paste its prompt; you open its
https://api.webly.ai/v1/mcplink and the answer contains your key, once: save it in your secure credentials, never the chat (agent.md, Phase 3b). Expired? Give them the renew link it returns. To claim without a browser, print/connect/<code>(nowebly claim-link) for them to open on their phone: keys can't claim.--openalso recordsconnect, so the claim is finished by whichever session loads the tools first. Itspending: claimfield says what to do.next
- Claude Code:
- Use MCP if it's already callable; otherwise claim in the browser now.
Check whether a Webly tool is callable in this session (search deferred tools). If yes, go to step 4. If not (the usual case right after a first
create_claim_code), don't ask the person to reload or restart. Runconnectstraight away. It openswebly claim-link --openwith the token in the URL fragment (never sent to a server, never printed). Ask them to sign in and click Claim website. If the page shows the wrong Google account, Not you? Use a different account switches it. Opening the page or signing in alone does not claim the site. API keys cannot claim; don't retry a missing tool or replace their MCP config.app.webly.ai/claim - Confirm the browser claim. After they say they're done, run
: only
webly doctorconfirms the token was claimed (doctor removes the spent token and clears pending). If it is still unclaimed, the browser was closed, or status is unavailable, keep the token and pending step so they can retry. Never runsiteError.reason: credential_consumedmerely because the browser opened. MCP loads by itself next session (Codex:forgetfirst). Stop here; steps 4-5 are the MCP path.codex mcp login webly - When the tools are loaded:
- Claude Code: call the server's tool (
authenticateormcp__plugin_webly_webly__authenticate; load it with tool search if it's deferred). It returns an authorization URL. Open it for them (mcp__webly__authenticateon macOS,open "<url>"on Linux,xdg-openon Windows) and print it as a fallback. The callback completes by itself.start - Other hosts: call ; the 401 starts the host's own sign-in.
whoami - First they choose where you can work: All my workspaces (the default; tell them to keep it unless they want you limited to one) or a single workspace. Either way you get only what their role allows in each.
- The consent page offers ,
webly:contentandwebly:editand defaults towebly:admin, which creating sites needs. Tell them to keep it unless they only want you editing existing sites (webly:admin).webly:edit
- Claude Code: call the server's
- Claim: call , then run
create_claim_code. The helper sends the code and the saved token to Webly itself, so you never read the token, and on success it deletes the token and the pending step. Tell them which site moved and give the printedwebly claim <code>. A retry by the same person is safe (dashboardUrl). The code lasts 10 minutes; get a new one if it expired (alreadyClaimed: true).claim_code_invalid
After the claim: with billing off, or on a paid plan with a free slot, the site
is permanent. On the Free plan with billing on it's held: editable until the
original 24-hour mark, then offline until they upgrade (). Say
which one the response shows (); don't promise permanence.
402 claim_heldwebsite.claimHeld当用户想保留站点、创建账户或登录,或者 仅为 时,执行此流程。先说明将会发生什么:浏览器页面会打开,用户登录(Google)并批准访问,随后站点会移入其自己的工作区。
next.actions["claim"]- 在此主机上注册 MCP 服务器,使用用户级作用域,使其能够跨新会话和其他文件夹继续使用:
- Claude Code:运行 (安装 Webly 插件,或回退到
webly connect claude)。claude mcp add --scope user - Codex:运行 。
webly connect codex - 其他主机:遵循 https://webly.ai/agent.md 的第 2 阶段。
- 当前环境没有浏览器(例如手机或 Muse 等托管式助手):如果应用有自己的连接器界面,让用户在其中添加 并通过它登录。如果它只接受 API key:让用户打开控制面板中的 Connect agent 并粘贴其中的提示;你打开其
https://api.webly.ai/v1/mcp链接,答复中会包含一次性的 key:将其保存在安全凭据中,绝不要放在聊天里(参见 agent.md 第 3b 阶段)。如果已过期,提供其返回的续期链接。要在没有浏览器的情况下认领,运行/connect/<code>(不要加webly claim-link),让用户用手机打开:API key 无法认领。--open还会记录connect,因此最先加载工具的会话将完成认领。其pending: claim字段会说明应执行的操作。next
- Claude Code:运行
- 如果 MCP 已可调用,则使用 MCP;否则立即在浏览器中认领。 检查当前会话中能否调用 Webly 的 工具(搜索延迟加载的工具)。如果可以,进入第 4 步。如果不可以(首次执行
create_claim_code后通常如此),不要要求用户重新加载或重启。直接运行connect。它会打开webly claim-link --open,Token 位于 URL fragment 中(绝不会发送到服务器,也绝不会被打印)。请用户登录并点击 Claim website。如果页面显示了错误的 Google 账户,Not you? Use a different account 可切换账户。 仅打开页面或登录并不能认领站点。API key 无法认领;不要反复尝试查找缺失的工具,也不要替换其 MCP 配置。app.webly.ai/claim - 确认浏览器中的认领结果。 用户表示完成后,运行 :只有
webly doctor才能确认 Token 已被认领(doctor 会删除已用完的 Token 并清除 pending)。如果仍未认领、浏览器被关闭或无法获取状态,则保留 Token 和 pending 步骤,以便用户重试。绝不要仅仅因为打开过浏览器就运行siteError.reason: credential_consumed。MCP 会在下次会话自行加载(Codex:先运行forget)。到此为止;第 4~5 步属于 MCP 路径。codex mcp login webly - 工具已加载时:
- Claude Code:调用服务器的 工具(
authenticate或mcp__plugin_webly_webly__authenticate;如果它是延迟加载工具,则通过工具搜索加载)。它会返回授权 URL。为用户打开该 URL(macOS 使用mcp__webly__authenticate,Linux 使用open "<url>",Windows 使用xdg-open),并将其打印出来作为后备方式。回调会自动完成。start - 其他主机:调用 ;401 会启动主机自身的登录流程。
whoami - 首先由用户选择允许你工作的位置:All my workspaces(默认选项;除非用户希望仅限于一个工作区,否则应保留此选项),或选择单个工作区。无论哪种方式,你只能获得其角色在各工作区中允许的权限。
- 同意页面会提供 、
webly:content和webly:edit,默认选择webly:admin,创建站点需要此权限。除非用户只希望你编辑现有站点(webly:admin),否则应告知用户保留默认权限。webly:edit
- Claude Code:调用服务器的
- **认领:**调用 ,然后运行
create_claim_code。辅助工具会将代码和已保存的 Token 直接发送给 Webly,因此你无需读取 Token;成功后,它会删除 Token 和 pending 步骤。告知用户哪个站点已移入账户,并提供打印出的webly claim <code>。同一人重试是安全的(dashboardUrl)。代码有效期为 10 分钟;如果已过期(alreadyClaimed: true),请获取新代码。claim_code_invalid
认领后:如果未启用付费功能,或使用付费套餐且仍有免费名额,站点将永久保留。在启用付费功能的 Free 套餐上,站点会处于保留状态:最初 24 小时标记前仍可编辑,之后会下线,直至用户升级()。应说明响应中显示的是哪一种状态();不要承诺永久保留。
402 claim_heldwebsite.claimHeldWorking over MCP (signed in)
通过 MCP 工作(已登录)
Call and first, and continue an existing site (the
folder's linked one, if shows a link) rather than creating a second
one. After for a project on disk, run in the project folder so later sessions find it.
whoamilist_websitesdoctorcreate_websitewebly link <websiteId>Workspaces. → means this connection works in
every workspace your person belongs to. names them with their
role in each. A site id works in any workspace, so site tools need nothing more.
Tools that act on a whole workspace (, ,
…) take its slug as . If they have several, ask which
one before creating a site, and on ask; never guess.
means the connection is limited to one workspace. If they
need another ( shows ), they reconnect: in
Claude Code → webly → re-authenticate, and pick All my workspaces.
When the owner asks about the "Made with Webly" badge: Hobby and Pro sites go
without it unless they opt in. turns it on or off for every
site, new ones included; with sets one site, and that
always wins ( follows the workspace again). Free always shows it: on
, give them .
whoamiworkspaces: "all"list_workspacescreate_websitelist_websitesget_billingworkspace400 workspace_requiredworkspaces: "one"list_workspacescanUse: false/mcpset_workspace_badgeupdate_websiteshowBadgenull402 plan_requireddetails.upgradeUrlWhen the owner asks for a team workspace photo, takes a
public image or the file as (PNG, JPEG, WebP or GIF, 2 MB);
clears it. A personal workspace shows the person's own
profile photo, which they change in the dashboard.
set_workspace_photourlbase64remove_workspace_photoREST without MCP (scripts, hosts with no MCP): opens the
browser to the same sign-in and workspace choice, then (or any path, for a JSON body) calls the API
as the person. The token stays in the helper; never ask for it or print it.
revokes it.
webly loginwebly api GET /websites --workspace <slug>/v1--datawebly logout- Framework sites (typed React, the default): /
put_source_file(no lease needed). Static sites:str_replaceandsearch_sourcewithstr_replacefor small text edits;{ path, edits }only for a few small whole files, withdeploy_filesset to the versionexpectedHeadVersionreturned (read_source_fileswithout it). If someone wrote since, your change is merged onto theirs (400 base_requiredin the response);mergedmeans you both changed the same lines: tell the person whose change it is (409 merge_conflict), re-read, re-apply, resend.theirsByonly when the person asks. For a folder on disk (a build'sforce, anything with bundles or images), calldist/with the site's id and straight away runbegin_deploy(the ticket lasts 5 minutes and one deploy; never echo it back to the person). Every write (deploy,webly deploy <folder> --ticket <ticket> --message "<subject>" --body "<what changed and why>",put_source_file, deletes) takes a message, written like a git commit message: a subject line of what changed, about 50 characters, imperative ("Show yearly plans first on Pricing"), then a blank line and a body in plain words for the site owner: what changed and why, which pages, anything to check. The site's History shows the subject, and the body when someone opens the version, so the people sharing the site can tell your version from everyone else's. Skip the body only for a trivial change. The helper builds the project if needed, uploads it (up to 2000 files; images and video over 1 MiB go straight to storage, up to the plan's per-file limit), links the folder and prints the draft URL. Never put a bundle's bytes in a tool call or an API key in a shell command. If it stops withstr_replace, the deploy would undo changes made to the site since this folder last deployed it (the paths are listed): read them withwould_revert, merge them into the local project, rebuild, and deploy again with a new ticket; or tell the person what would be undone and, only with their OK, deploy with a new ticket andread_source_files. If it stops with--force, the site has files the folder doesn't (listed): add them to the folder, or only with the person's OK deploy withwould_remove. A clone of the site's git mirror (--force) deploys the same way: the helper keeps the images listed inget_git_accessand takes its base from the git history; when it says the clone is behind,.webly/media.jsonfirst. If the result hasgit pull, tell the person the site had changed elsewhere and earlier versions are kept. Each write makes a new draft version; the live site doesn't change. Writes are Prettier-formatted, so copyreplacedtext fromstr_replace.read_source_file - runs the quality gate (lint, typecheck, bundle, render). Fix every diagnostic.
check_head - Show the person the draft URL (, not public, not indexed).
https://draft--{subdomain}.webly.site - only after they say yes to that exact site and version. It waits for a running build;
publish_websitewith202means Webly is still building (not a failure): publish that version again in about 15 seconds, and tell the person it's still building.status: "building"/rollback_websiteif something is wrong live.unpublish_website
New sites get a short generated address (); don't pass
unless the person asks for a specific one (up to 56 characters; a taken
one is a , not a variant). On Free each site can change its webly.site
address once (; Hobby and Pro unlimited), so run
first and rename only to the exact name the person asked for.
On Hobby and Pro the workspace also has a namespace
and each published site answers at
(, the label from the site's name, unique only
within the namespace). Share : the primary address (a custom domain
when one is live, else the vanity, else the flat address, unless the owner chose a
webly.site address as primary); lists every
address the site answers at.
changes a site's label; / /
handle the namespace. Renames are scarce (Hobby 5, Pro 10, lifetime), so rename
only to the exact name the person asked for, and tell them old vanity addresses
stop working. The new addresses need about a minute for HTTPS: check
until is , then tell them they're live.
On , give them .
zen-wreath-y76r.webly.sitesubdomain409rename_subdomaincheck_subdomainurls.vanityhttps://fall-out-boy.kev.webly.siteurls.primaryurls.allrename_vanityget_namespacecheck_namespacerename_namespaceget_namespacecertificateactive402details.upgradeUrlA team workspace has one owner ( → ). Ownership of the workspace
itself (not moving a website) changes only in the dashboard's People page, by the
owner; you can't do it, so send them there. The card and subscription stay with
the workspace.
whoamiownerNew content, including blog posts and repeating content (products, team, FAQs),
goes in the site's source files; don't create collections or blog posts, the CMS
is deprecated. A site that already reads collections or a managed blog keeps
using those tools for that content until the owner asks to move it. Forms
post with , never , and never ask for card
numbers, passwords, ID numbers or health details (submissions are plain text).
If the owner wants an email per submission, turn it on with
(off by default). Emails go out on Hobby and Pro;
on Free the choice is saved: give the owner . Name the email
field so replying to an alert answers the visitor. Each workspace takes
1,000 submissions a month on Free (10,000 Hobby, 100,000 Pro); past that they're
held: hidden, no alerts, and deleted at the monthly reset unless the owner
upgrades. When shows , tell the owner and give them
.
Details are in llms.txt.
formAction('name')mailto:set_form_notificationsupgradeUrlemaillist_formsquota.heldupgradeUrl403details.capability401invalid_token/webly.css<link>str_replacebaseVersionversionput_source_filedelete_source_file409 merge_conflictwait_for_changesversionalsoPublishesalsoPublisheslist_changesdiff_source_filepreview_versionGit: gives a read-only clone of a signed-in site's source and
history (one commit per version). Its token lasts one minute, so run its
(or inside the clone) straight away as
, never with the token in the URL. Check with
before asking again. Never ; write through the tools.
get_git_accessclonepullWEBLY_GIT_TOKEN=<token> <clone>headVersionget_websitegit push首先调用 和 ,并继续使用现有站点(如果 显示文件夹已关联,则使用该关联站点),而不是创建第二个站点。通过 为磁盘上的项目创建站点后,在项目文件夹中运行 ,以便后续会话找到它。
whoamilist_websitesdoctorcreate_websitewebly link <websiteId>工作区。 → 表示此连接可用于用户所属的每个工作区。 会列出工作区及其在每个工作区中的角色。站点 ID 在任何工作区中都有效,因此站点工具不需要其他信息。对整个工作区执行操作的工具(、、 等)需要将其 slug 作为 参数。如果用户有多个工作区,创建站点前应询问使用哪一个;遇到 时也应询问,绝不猜测。 表示连接仅限于一个工作区。如果用户需要其他工作区( 显示 ),则需要重新连接:在 Claude Code 中进入 → webly → re-authenticate,并选择 All my workspaces。当网站所有者询问“Made with Webly”徽标时:Hobby 和 Pro 站点默认不显示该徽标,除非用户主动选择启用。 可为所有站点(包括新站点)统一开启或关闭; 使用 可单独设置某个站点,且单独设置始终优先( 表示重新遵循工作区设置)。Free 套餐始终显示该徽标:遇到 时,提供 。
whoamiworkspaces: "all"list_workspacescreate_websitelist_websitesget_billingworkspace400 workspace_requiredworkspaces: "one"list_workspacescanUse: false/mcpset_workspace_badgeupdate_websiteshowBadgenull402 plan_requireddetails.upgradeUrl当网站所有者需要团队工作区照片时, 可接受公开图片的 ,也可接受以 提供的文件(PNG、JPEG、WebP 或 GIF,最大 2 MB); 会清除照片。个人工作区显示用户自己的个人资料照片,用户可以在控制面板中修改。
set_workspace_photourlbase64remove_workspace_photo不使用 MCP 的 REST(脚本或没有 MCP 的主机): 会打开浏览器,进入同一个登录及工作区选择流程;然后 (或任何 路径,使用 提供 JSON 请求体)会以用户身份调用 API。Token 保留在辅助工具中;绝不要索取或打印它。 会撤销它。
webly loginwebly api GET /websites --workspace <slug>/v1--datawebly logout- 框架站点(类型化 React,默认):使用 /
put_source_file(无需租约)。静态站点:小规模文本编辑使用str_replace和带search_source的{ path, edits };str_replace仅用于少量完整的小文件,并将deploy_files设置为expectedHeadVersion返回的版本(缺少该值会返回read_source_files)。如果期间有人进行了写入,你的更改会与对方的更改合并(响应中显示400 base_required);merged表示双方修改了相同行:告知用户其变更是哪一方的(409 merge_conflict),重新读取、重新应用并重新发送。只有用户明确要求时才使用theirsBy。对于磁盘上的文件夹(构建生成的force、任何包含 bundle 或图片的目录),使用站点 ID 调用dist/,并立即运行begin_deploy(工单有效期为 5 分钟且只能使用一次;绝不要将工单回显给用户)。每次写入(部署、webly deploy <folder> --ticket <ticket> --message "<subject>" --body "<what changed and why>"、put_source_file、删除)都需要提供一条消息,写法类似 Git 提交消息:先写约 50 个字符的祈使句主题(例如“在定价页优先显示年度套餐”),然后空一行,再用通俗语言写正文,说明改了什么、为什么改、涉及哪些页面以及需要检查什么。站点的 History 会显示主题,用户打开某个版本时还会看到正文,因此共享网站的人可以区分你的版本与其他人的版本。仅在进行微小更改时可以省略正文。辅助工具会按需构建项目并上传(最多 2000 个文件;超过 1 MiB 的图片和视频会直接进入存储,受套餐单文件上限约束),关联文件夹并打印草稿 URL。绝不要在工具调用中放入 bundle 的字节,也不要在 shell 命令中加入 API key。如果命令以str_replace停止,说明部署会撤销自此文件夹上次部署以来对站点所做的更改(路径会列在输出中):使用would_revert读取这些更改,将它们合并到本地项目中,重新构建,再使用新工单部署;或者告知用户哪些内容会被撤销,并在获得用户同意后使用新工单和read_source_files部署。如果命令以--force停止,说明站点存在文件夹中没有的文件(会列在输出中):将这些文件加入文件夹,或者仅在用户同意后使用would_remove部署。已登录站点的 Git 镜像克隆(--force)也使用相同方式部署:辅助工具会保留get_git_access中列出的图片,并从 Git 历史中采用其基线;当它提示克隆落后时,先运行.webly/media.json。如果结果中包含git pull,告知用户该站点在其他位置发生过更改,并且较早版本仍然保留。每次写入都会创建新的草稿版本;线上站点不会改变。写入内容会按 Prettier 格式化,因此应从replaced复制read_source_file所需文本。str_replace - 会运行质量门禁(lint、类型检查、bundle、渲染)。修复所有诊断信息。
check_head - 向用户展示草稿 URL(;该地址不公开,也不会被索引)。
https://draft--{subdomain}.webly.site - 只有当用户明确同意该确切站点和版本后,才调用 。它会等待正在进行的构建;返回
publish_website且202表示 Webly 仍在构建(并非失败):约 15 秒后再次发布该版本,并告知用户它仍在构建。如果线上内容有问题,则使用status: "building"/rollback_website。unpublish_website
新站点会获得一个简短生成的地址();除非用户要求指定地址,否则不要传入 (最长 56 个字符;地址已被占用时返回 ,不会提供变体)。在 Free 套餐上,每个站点的 webly.site 地址可以更改一次(使用 ;Hobby 和 Pro 不限次数),因此应先运行 ,且只在用户要求的确切名称时执行重命名。在 Hobby 和 Pro 套餐上,工作区还拥有命名空间,每个已发布站点都可通过 访问(例如 ;标签来自站点名称,仅在命名空间内唯一)。应分享 :主地址(自定义域名已上线时为自定义域名,否则为 vanity 地址,否则为扁平地址,除非所有者选择了某个 webly.site 地址作为主地址); 列出站点可响应的全部地址。 更改站点标签; / / 用于管理命名空间。重命名次数很少(Hobby 5 次、Pro 10 次、终身套餐同样有限),因此只在用户要求的确切名称时执行,并告知用户旧的 vanity 地址将停止工作。新地址大约需要一分钟才能启用 HTTPS:检查 ,直至 为 ,然后告知用户地址已上线。遇到 时,提供 。
zen-wreath-y76r.webly.sitesubdomain409rename_subdomaincheck_subdomainurls.vanityhttps://fall-out-boy.kev.webly.siteurls.primaryurls.allrename_vanityget_namespacecheck_namespacerename_namespaceget_namespacecertificateactive402details.upgradeUrl团队工作区只有一位所有者( → )。工作区本身的所有权转移(而非移动网站)只能由所有者通过控制面板的 People 页面完成;你无法执行,因此应引导用户前往该页面。工作区卡片和订阅仍归工作区所有。
whoamiowner新内容,包括博客文章和重复性内容(产品、团队、FAQ),应放入站点的源文件中;不要创建 collections 或博客文章,CMS 已弃用。仍在读取 collections 或使用托管博客的站点,应继续对这些内容使用相应工具,直到所有者要求迁移。表单使用 提交,绝不使用 ,也绝不要索取卡号、密码、身份证号或健康信息(提交内容为纯文本)。如果所有者希望每次提交都发送邮件,则开启 (默认关闭)。邮件功能仅在 Hobby 和 Pro 套餐上可用;在 Free 套餐上会保存该选择:向所有者提供 。将电子邮件字段命名为 ,这样回复提醒时可直接答复访客。每个工作区每月可接收 1,000 次提交(Free;Hobby 为 10,000 次,Pro 为 100,000 次);超出后,提交会被保留:隐藏、不发送提醒,并在月度重置时删除,除非所有者升级。当 显示 时,告知所有者并提供 。
formAction('name')mailto:set_form_notificationsupgradeUrlemaillist_formsquota.heldupgradeUrl详细信息见 llms.txt。
403details.capability401invalid_token/webly.css<link><link>str_replaceput_source_filedelete_source_filebaseVersionversion409 merge_conflictwait_for_changesversionalsoPublishesalsoPublisheslist_changesdiff_source_filepreview_versionGit: 会提供已登录站点源文件和历史的只读克隆(每个版本对应一次提交)。其 Token 仅一分钟有效,因此应立即以 的形式运行 (或在克隆中运行 ),绝不要把 Token 放入 URL。再次询问用户之前,使用 检查 。绝不要运行 ;应通过工具进行写入。
get_git_accessWEBLY_GIT_TOKEN=<token> <clone>clonepullget_websiteheadVersiongit pushTurn a static site into a React site (or back)
将静态站点转换为 React 站点(或反向转换)
Create a new site of the other kind and port the pages yourself: components,
and React state for what scripts did (or plain HTML and small scripts the
other way). Leave the original untouched until your person is happy with the
new one.
创建另一个类型的新站点,并自行迁移页面:迁移组件,以及原本由脚本实现的 React state(反向转换时则改为纯 HTML 和小型脚本)。在用户对新站点满意之前,保持原站点不变。
Duplicate or move a site
复制或移动站点
Hobby and Pro, in both workspaces ( otherwise: give the
person ). copies a site 1:1: every
version, media, CMS, form submissions and settings, passwords; it is published
if the original is ( means a bigger site is still
building and goes live by itself: tell your person it is still working), gets its own addresses, and custom domains stay with the
original. moves a site into another workspace your person
admins, at once: everything comes along, including its webly.site address and
custom domains, except form email alerts: their recipients are cleared, so set
them again there (). Its vanity address changes to the
new workspace's and links to the old one stop working, so confirm first and
share . lists the workspaces your person
admins and whether the site fits (: , ,
); ask which one. Both need a signed-in person (OAuth) to reach
another workspace. To give it to someone else, ask for their email and call
: they're emailed and it waits on their
Transfers page, where they accept or decline it (or their agent:
, ,
). There are no transfer codes.
402 plan_requireddetails.upgradeUrlduplicate_websitepublish.status: "building"move_websiteset_form_notificationsurls.primarylist_move_destinationsreasonsite_limitstorage_limitplan_requiredstart_website_transfer { toEmail }list_incoming_transfersaccept_website_transfer { transferId }decline_website_transferHobby 和 Pro 套餐支持此操作,在两个工作区中均可使用(否则返回 :向用户提供 )。 会以 1:1 比例复制站点:所有版本、媒体、CMS、表单提交和设置、密码都会复制;如果原站点已发布,副本也会发布( 表示较大的站点仍在构建,并会自动上线:告知用户它仍在正常工作),副本会获得自己的地址,而自定义域名仍归原站点所有。 会将站点立即移入用户拥有管理员权限的另一个工作区:所有内容都会一并移动,包括其 webly.site 地址和自定义域名,但表单邮件提醒的收件人会被清空,需要在目标工作区重新设置()。其 vanity 地址会变为新工作区的地址,旧地址将停止工作,因此应先确认并分享 。 会列出用户拥有管理员权限的工作区,以及站点是否适合移动(:、、);如有多个选项,应询问用户。两个操作都需要已登录用户(OAuth)才能访问另一个工作区。若要将站点交给他人,询问其电子邮件地址并调用 :系统会向该地址发送邮件,用户需在其 Transfers 页面中接受或拒绝(也可由其 Agent 调用 、 或 )。不存在转移代码。
402 plan_requireddetails.upgradeUrlduplicate_websitepublish.status: "building"move_websiteset_form_notificationsurls.primarylist_move_destinationsreasonsite_limitstorage_limitplan_requiredstart_website_transfer { toEmail }list_incoming_transfersaccept_website_transfer { transferId }decline_website_transferCustom domains
自定义域名
Needs a signed-in account (MCP) on Hobby (5 domains) or Pro (25). On Free,
is : give the owner .
"Connect example.com to my site":
add_domain402 domain_limitdetails.upgradeUrl- Pick the site with . Never guess it from the hostname.
list_websites - For a bare domain, ask first:
- (recommended). It works with every DNS provider. The person then forwards
www.example.comto it at their registrar, which is free and takes a minute, but usually http-only (seeexample.combelow).apexForward - directly. Only if their DNS provider supports CNAME flattening or ALIAS at the root (Cloudflare, Namecheap, Porkbun…). For a subdomain such as
example.com, just add it.shop.example.com
- Call , then
add_domain. If it returns aconnect_domain, give the person that link first: they sign in at their DNS provider (Cloudflare for any hostname; GoDaddy and others for www once they support it) and approve, and the records are set for them. They open it in the browser where they use Webly. With Cloudflare (url) Webly also forwards the bare domain of a www domain if they allow redirects on Cloudflare's screen and nothing is served there yet (ifmethod: "oauth"isn't ok after verifying, give it to them). If records of theirs are in the way, nothing is replaced until they approve the exact changes shown after signing in; approving switches the hostname at once, so for a site live today offer the manual records (TXT first) to move without downtime. When youapexForwardand the response hasremove_domain, ask whether to delete the records Webly added too, and if yes give themdnsCleanup. Otherwise (name.com, Squarespace, Google and most other providers), or as the fallback, give every requireddnsCleanup.urlentry exactly as returned, as a Type / Name / Value table: the CNAME and thedns.recordsTXT records. The TXT issues the HTTPS (SSL) certificate and is always needed; without it HTTPS can take hours. If it isn't in_acme-challengeyet, calldns.recordsa few seconds later and give it then. For a www domain, also give theverify_domainforwarding rule (URL forwarding at their registrar makes the bare domain work too, usually over plain http only; forapexForwardlinks, a provider with ALIAS can add the bare domain as a second domain on the site: TXT records first, the ALIAS once it's active). Check it withhttps://(verify_domain,apexForward.ok), not in a browser: Chrome quietly retries a bare domain withapexForward.https. Provider tips:www- Type only the Name shown (, or
wwwfor the root). GoDaddy and Namecheap add the domain themselves.@ - On Cloudflare, set the record to "DNS only" (grey cloud).
- Moving a domain that is live elsewhere? Keep its current record and add
only the TXT records (the certificate ones and the optional ownership
one). Then:
- Poll until the domain is
verify_domain. That means ownership is verified and the certificate is deployed. Step 2 isn't enough.activecan't show this yet, because the old host still answers.domain-watch - Ask the person to switch the CNAME.
- Run to confirm traffic now reaches Webly.
webly domain-watch <hostname>
- Poll
- Type only the Name shown (
- Start watching in the background right away. Don't wait for the person to
say they're done:
as a background task. It prints
webly domain-watch <hostname>,dnsandhttpsas each happens and exits once the domain is live. Without a shell, callliveeveryverify_domain. Every domain response hasnext.checkAgainSecondswith the step (1 DNS record, 2 HTTPS certificate, 3 live) and what to do. Tell the person when each step completes.next - Once live: confirm loads. The first domain to go live on a site becomes its primary address by itself (
https://<hostname>), unless a primary was already chosen (another domain, or a webly.site address picked withisPrimary); every other address redirects to the primary. If another domain is already primary, offerset_primary_addressto switch, orset_primary_domainif they want a webly.site address to be the primary (their custom domains then redirect to it; say so). If it'sset_primary_address, relayfailed, have them fix the record, then callfailureMessageagain. If the domain is active but their browser says the server "could not be found" right after removing and re-adding it, their DNS cached the gap: it clears within the zone's negative-cache time (30 minutes by default on Cloudflare), or sooner if they flush their DNS cache and the stale answer was cached on their device (not by their router or ISP).verify_domain
Adding a domain publishes nothing and takes nothing offline. The
address keeps working.
*.webly.site此功能需要已通过 MCP 登录的账户,并需要 Hobby(5 个域名)或 Pro(25 个域名)套餐。在 Free 套餐上, 会返回 :向所有者提供 。
add_domain402 domain_limitdetails.upgradeUrl“将 example.com 连接到我的站点”:
- 使用 选择站点。绝不要根据主机名猜测站点。
list_websites - 对于裸域名,应先询问:
- (推荐)。 它适用于所有 DNS 提供商。用户随后可在注册商处将
www.example.com转发到该地址;此操作免费且约需一分钟,但通常仅支持 HTTP(参见下文的example.com)。apexForward - 直接使用 。 仅当其 DNS 提供商支持根域名的 CNAME flattening 或 ALIAS 时使用(Cloudflare、Namecheap、Porkbun 等)。 对于
example.com等子域名,直接添加即可。shop.example.com
- 调用 ,然后调用
add_domain。如果返回connect_domain,应先把该链接提供给用户:用户需在其 DNS 提供商处登录(任何主机名均可使用 Cloudflare;GoDaddy 等提供商在支持 www 后也可使用),并批准相关设置,记录将自动完成。随后,用户需在使用 Webly 的浏览器中打开该链接。对于 Cloudflare(url),如果用户允许在 Cloudflare 页面重定向,且该裸域名尚未提供任何内容,Webly 还会转发 www 域名的裸域名(如果验证后method: "oauth"状态不正确,则将链接提供给用户)。如果用户自己的记录造成冲突,在用户登录并批准登录后显示的确切更改之前,不会替换任何内容;批准后会立即切换主机名。因此,对于当前已上线的站点,应提供手动记录(先配置 TXT)以实现无停机迁移。当调用apexForward且响应包含remove_domain时,询问用户是否也要删除 Webly 添加的记录;如果同意,则提供dnsCleanup。 其他提供商(name.com、Squarespace、Google 和大多数其他提供商)或作为后备方案时,应将每个必需的dnsCleanup.url条目完全按返回结果提供,并整理为 Type / Name / Value 表:其中必须同时包含 CNAME 和dns.recordsTXT 记录。TXT 记录用于签发 HTTPS(SSL)证书,始终必需;缺少该记录时,HTTPS 可能需要数小时才能完成。如果_acme-challenge中暂时还没有这些内容,请在几秒后调用dns.records,届时再提供给用户。对于 www 域名,还应提供verify_domain转发规则(在注册商处进行 URL 转发通常也能让裸域名生效,但一般仅支持纯 HTTP;对于apexForward链接,支持 ALIAS 的提供商可以将裸域名作为站点的第二个域名添加:先配置 TXT 记录,ALIAS 生效后再添加)。使用https://检查verify_domain和apexForward.ok,不要只在浏览器中检查:Chrome 会在静默重试 www 域名的裸域名时忽略此问题。提供商提示:apexForward.https- Type 只填写显示的 Name(,根域名则为
www)。GoDaddy 和 Namecheap 会自行添加域名。@ - 在 Cloudflare 中,将记录设置为“DNS only”(灰色云朵)。
- 正在迁移已在其他位置上线的域名?保留其当前记录,只添加 TXT 记录(证书记录和可选的所有权记录)。然后:
- 轮询 ,直至域名为
verify_domain。这意味着所有权已验证且证书已部署,仅完成第 2 步并不足够。由于旧主机仍在响应,此时active还无法显示结果。domain-watch - 请用户切换 CNAME。
- 运行 ,确认流量现已到达 Webly。
webly domain-watch <hostname>
- 轮询
- Type 只填写显示的 Name(
- 立即在后台开始监视,不要等待用户表示已经完成:
将 作为后台任务运行。它会依次打印
webly domain-watch <hostname>、dns和https,并在域名上线后退出。如果没有 shell,则每隔live调用一次next.checkAgainSeconds。每个域名响应都包含verify_domain,其中会指明步骤(1 DNS 记录、2 HTTPS 证书、3 上线)和应执行的操作。每一步完成时都应告知用户。next - 上线后:确认 能够加载。站点上第一个成功上线的域名会自动成为主地址(
https://<hostname>),除非已经选定主地址(另一个域名,或通过isPrimary选定的 webly.site 地址);其他所有地址都会重定向到主地址。如果另一个域名已经是主地址,可提供set_primary_address进行切换;如果用户希望 webly.site 地址成为主地址,则使用set_primary_domain(此时自定义域名会重定向到该地址,应明确告知)。如果状态为set_primary_address,传达failed,请用户修复记录,然后再次调用failureMessage。如果域名处于 active 状态,但用户在移除并重新添加域名后,浏览器却提示“找不到服务器”,则其 DNS 缓存了这段空档:在 Cloudflare 中,默认会在该区域的负缓存时间后清除(通常为 30 分钟);如果过期答案缓存在用户设备上,清除其 DNS 缓存后可能更快恢复(问题不在路由器或 ISP 上)。verify_domain
添加域名不会发布任何内容,也不会使任何内容下线。
地址会继续有效。
*.webly.siteRules
规则
- The token is a secret that proves ownership. Never print it, log it, commit
it, or put it in a URL other than through .
webly claim-link - Only state URLs, sites and counts that a command or tool actually returned.
- Don't set up MCP for someone who only wants a quick link; offer it when they want to keep or grow the site.
- One anonymous site per machine. If they want a second site kept, claim the
first, then create the second over MCP and it.
webly link - A folder belongs to the site in its . Never deploy it to, or claim, a different site unless the person asks.
.webly/project.json
- Token 是证明所有权的机密信息。绝不打印、记录或提交它,也不要将其放入 以外的 URL。
webly claim-link - 只陈述命令或工具实际返回的 URL、站点和数量。
- 不要为只需要临时链接的用户配置 MCP;当用户想保留或扩展站点时再提供。
- 每台机器只能拥有一个匿名站点。如果用户想永久保留第二个站点,应先认领第一个,再通过 MCP 创建第二个并运行 。
webly link - 文件夹归属于其 中指定的站点。除非用户明确要求,否则绝不将该文件夹部署到其他站点,也绝不认领其他站点。
.webly/project.json