> ## Documentation Index > Fetch the complete documentation index at: https://autclaw.shop/llms.txt > Use this file to discover all available pages before exploring further. # JavaScript Sender message capabilities > Use JavaScript Sender methods to read users and chats, match parameters, reply with text or media, send mixed messages, and recall messages. `Sender` binds to the current runtime context. For a message trigger, it represents the current IM, account, chat, and user. For a route trigger, it represents the current HTTP request. A cron / fake context does not point to a real IM chat. ```js theme={null} const { Sender, getSenderID } = require('./middleware.js') const sender = new Sender(getSenderID()) ``` ## Read message context These methods take no parameters and return fields from the current runtime context. When a field does not exist, they usually return an empty string, empty array, or `false`. | Method | Purpose | Parameters | Returns | | ------------------------------ | ------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------------- | ------------------- | | `getImtype()` | Get the current IM type. Cron runs return `fake`. | None | `Promise` | | `getUserID()` | Get the current user ID. | None | `Promise` | | `getUserName()` | Get the current user display name. | None | `Promise` | | `getUserAvatarUrl()` | Get the current user avatar URL. | None | `Promise` | | `getChatID()` | Get the current chat ID. For groups this is the group ID; for private chats it is usually empty. | None | `Promise` | | `getChatName()` | Get the current chat name. | None | `Promise` | | `getGroupName()` | Get the group name. Same as `getChatName()`. | None | `Promise` | | `isAdmin()` | Check whether the current user is an administrator. | None | `Promise` | | `isAI()` | Check whether the current message came from an AI trigger context. | None | `Promise` | | `getAIParam()` | Get the AI parameter. | None | `Promise` | | `getMessage()` | Get the current message text. | None | `Promise` | | `getMessageID()` | Get the current message ID. | None | `Promise` | | `getHistoryMessageIDs(number)` | Get recent message IDs. | `number`: required, expected count. | `Promise` | | `param(index)` | Read a `rule` regex capture. | `index`: required, starts at `1`; out-of-range returns an empty string. | `Promise` | ```js theme={null} const user = await sender.getUserName() const message = await sender.getMessage() const firstMatch = await sender.param(1) await sender.reply(`${user} sent: ${message}\nParameter: ${firstMatch}`) ``` ## Reply with text and media `reply*` methods reply to the current runtime context: for a message trigger, they reply to the current chat; for a route trigger, `reply(...)` / `replyMarkdown(...)` writes the route response; the `fake` context for cron does not send a real IM reply. | Method | Purpose | Parameters | Returns | | ------------------------------ | ---------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------- | | `reply(text)` | Reply with plain text. | `text`: required text content. | `Promise` | | `replyMarkdown(markdown)` | Reply with Markdown text. Whether Markdown is rendered depends on adapter support. | `markdown`: required Markdown content. | `Promise` | | `replyImage(source, options?)` | Reply with an image. | `source`: required image URL, path, base64, or data URI; `options`: optional media options. | `Promise` | | `replyVoice(source, options?)` | Reply with voice/audio. | `source`: required voice URL or path; `options`: optional. | `Promise` | | `replyVideo(source, options?)` | Reply with a video. | `source`: required video URL or path; `options`: optional. | `Promise` | | `replyFile(source, options?)` | Reply with a file. | `source`: required file URL or path; `options`: optional. | `Promise` | | `replyMixed(items, options?)` | Reply with a structured mixed message. | `items`: required mixed items; `options`: optional and used as default options for media items. | `Promise` | | `returnValue(text)` | Set the runtime return value without directly sending an IM message. | `text`: required return text. | `Promise`, currently usually returns `"[]"` for compatibility. | ```js theme={null} await sender.reply('Plain text') await sender.replyMarkdown('## Result\n\n- Done') await sender.replyImage('https://example.com/a.png') await sender.replyVoice('./voice.wav') await sender.replyVideo('./demo.mp4') await sender.replyFile('./report.csv', { filename: 'report.csv', mimeType: 'text/csv' }) await sender.replyMixed([ { type: 'text', text: 'Result: ' }, { type: 'image', source: './result.png', name: 'result.png' }, ]) ``` Current-chat reply methods such as `replyImage(...)` do not require you to pass `imType`, group ID, or user ID. To specify a target, use [`push(...)`](/en/plugin-sdk/javascript/push) or `push*`. ## `MediaOptions` The second parameter of media reply methods is an optional object. File display name. You can also write `fileName` or `file_name`. The runtime normalizes it to media item `name`. MIME type. You can also write `mime_type`, such as `image/png`, `video/mp4`, or `text/csv`. Constraints passed to the media preparation flow. Use only when you need to limit output format or size. Extension options passed to the media preparation flow. Regular plugins usually do not need this. Reserved timeout field. Current direct media reply helpers use the built-in media timeout. ## `replyMixed(items, options?)` `items` is an array. Text items use `text`; media items use `source`, and can also use `file`, `url`, or `path` as source aliases. The runtime discards invalid mixed items. Mixed message array. Supports `text`, `markdown`, `image`, `voice`, `video`, and `file`. The runtime normalizes it before passing it to the adapter. Optional default media options. When a media item has no `name` / `mime_type`, values can be filled from here. ```js theme={null} await sender.replyMixed([ { type: 'text', text: 'Done: ' }, { type: 'image', source: './result.png', name: 'result.png' }, { type: 'file', source: './report.csv', name: 'report.csv', mime_type: 'text/csv' }, ]) ``` Supported items: | item | Required field | Optional fields | Description | | ----------------------------------- | -------------- | --------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------- | | `{ type: 'text', text: '...' }` | `text` | None | Send a plain text segment. | | `{ type: 'markdown', text: '...' }` | `text` | None | The helper passes it through. The current runtime normalization mainly accepts text and media items; adapter support depends on the implementation. | | `{ type: 'image', source: '...' }` | `source` | `name`, `mime_type`, `platform_payload` | Send an image. | | `{ type: 'voice', source: '...' }` | `source` | `name`, `mime_type`, `platform_payload` | Send voice/audio. | | `{ type: 'video', source: '...' }` | `source` | `name`, `mime_type`, `platform_payload` | Send a video. | | `{ type: 'file', source: '...' }` | `source` | `name`, `mime_type`, `platform_payload` | Send a file. | Example `SendReceipt`: ```json theme={null} { "ok": true, "message_id": "msg_mixed_123", "message_ids": ["msg_text_1", "msg_image_1"], "action": "sendMixed" } ``` ## Control flow | Method | Purpose | Parameters | Returns | | -------------------------- | --------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------- | ---------------------- | | `setContinue()` | Continue matching subsequent plugins. After a plugin is hit, later matches usually stop by default. | None | `Promise` | | `recallMessage(messageid)` | Recall the specified message. | `messageid`: required message ID. You can use `message_id` or `messageid`. | `Promise` | | `breakIn(content)` | Re-enter autClaw internal message handling with new text as a message event. | `content`: required new message text. | `Promise` | ```js theme={null} if (await sender.isAdmin()) { await sender.setContinue() } const receipt = await sender.reply('This message will be recalled in 5 seconds') setTimeout(() => { sender.recallMessage(receipt.message_id || receipt.messageid).catch(console.error) }, 5000) ``` ## Next steps See `listen(...)`, `input(...)`, and `waitPay(...)`. See `getMediaItems()` and `downloadAdapterFile(...)`.