discord-ticket-transcript

Archive any Discord channel (tickets, support rooms, chat logs) into a single self-contained HTML file with embedded structured JSON data.

• 📄 One file — Discord-style HTML viewer + machine-readable JSON, all in one
• 🖼 Survives CDN expiry — every image, attachment, emoji and avatar is inlined as base64
• 🌍 24 built-in languages — English, 한국어, 日本語, 中文 (简/繁), Español, Português (BR), Français, Deutsch, Italiano, Русский, Türkçe, Polski, Nederlands, العربية, हिन्दी, Bahasa Indonesia, Tiếng Việt, ไทย, Українська, Čeština, Svenska, Suomi, Ελληνικά
• 🧩 Every Discord surface — embeds, reactions, buttons, select menus, modals, stickers, threads, system messages, and Components V2 (Container / Section / TextDisplay / MediaGallery / Separator / File / Thumbnail)
• 🖱 Interactive viewer — right-click to copy message / user IDs & text, click avatar to see role/permission snapshot, click reactions to see who reacted, click reply to jump + highlight original message
• 🌓 Dark / light theme toggle
• 📦 JSON + plain-text export — download buttons baked into the HTML
• 👥 Captures viewer snapshot — which roles and members could see the channel at close time

Install

npm install discord-ticket-transcript discord.js
# or
yarn add discord-ticket-transcript discord.js

discord.js is a peer dependency (v14+). Install it alongside.

Node.js 16.9+ required.

Quick start

import { TicketTranscript, toAttachment } from 'discord-ticket-transcript'

// When a ticket is being closed:
const file = await TicketTranscript.create(channel, {
  ticketId: '0042',
  opener: openerUser,
  closedBy: interaction.user,
  category: 'Purchase inquiry',
  theme: 'toggle',
  languages: 'all',          // or 'ko' / ['ko','en','ja']
  defaultLanguage: 'en',
  meta: { orderId: 'A-12345', totalPrice: 39900 }
})

// Upload to a log channel
await logChannel.send({
  content: `Archived ticket: ${file.name}`,
  files: [toAttachment(file)]
})

// Or write to disk
import { writeFile } from 'node:fs/promises'
await writeFile(`./tickets/${file.name}`, file.buffer)

Open the generated .html in any browser. All assets are self-contained; no network access required.

API

TicketTranscript.create(channel, options?) → Promise<TranscriptFile>

Auto-fetches every message, reaction, attachment, embed, component, participant, and viewer from the given channel, then produces a single HTML file with embedded JSON data.

interface CreateOptions {
  ticketId?: string               // default: channel.id
  opener?: User | GuildMember
  closedBy?: User | GuildMember
  category?: string               // e.g. 'Purchase', 'Refund', 'Report'
  meta?: Record<string, unknown>  // free-form payload stored in the JSON

  theme?: 'dark' | 'light' | 'toggle'   // default: 'toggle'
  languages?: 'all' | LangCode | LangCode[]   // default: 'all'
  defaultLanguage?: LangCode                   // default: first in languages

  inlineAssets?: boolean          // base64-embed images/attachments (default true)
  maxAssetSize?: number           // skip embedding files larger than N bytes (default 8 MB)
  limit?: number                  // cap on messages to fetch (default unlimited)

  filename?: string               // default: ticket-<id>.html
  onProgress?: (info: { fetched: number }) => void
}

// Returns:
{ name: string, buffer: Buffer, attachment: Buffer }

toAttachment(file) → AttachmentBuilder

Wraps the result for channel.send({ files: [...] }).

TicketTranscript.extract(input) → Promise<TicketTranscriptData>

Pulls the embedded JSON back out of any previously generated HTML file. Accepts a file path, a Buffer, or an HTML string.

const data = await TicketTranscript.extract('./tickets/ticket-0042.html')
console.log(data.stats.totalMessages)
console.log(data.messages.length)
console.log(data.channel.viewers?.roles)

TicketTranscript.aggregate(dir) → Promise<{ count, totalMessages, tickets }>

Scans a directory of HTML transcripts and aggregates stats.

i18n helpers

import {
  SUPPORTED_LANGUAGE_CODES,    // ['en','ko','ja','zh-CN', ... 24 codes]
  LANGUAGE_DESCRIPTIONS,       // { ko: '한국어 (Korean)', ... }
  LOCALES,                     // full translation dictionaries
  DEFAULT_LANGUAGE,            // 'en'
  resolveLanguages
} from 'discord-ticket-transcript'

What's inside the generated HTML

| Section | Contents | |---|---| | <header> | Title, channel/guild, time range, language selector, theme toggle, JSON/TXT download | | <div class="tt-details"> | Collapsible summary cards, participants (users / bots split), channel viewers (roles / users / bots) | | <main class="tt-messages"> | Discord-styled message list | | <script id="ticket-data" type="application/json"> | Full TicketTranscriptData (see types) | | <script id="ticket-i18n" type="application/json"> | Translation dictionaries for chosen languages | | <script id="ticket-text" type="text/plain"> | Plain-text dump for one-click export |

The viewer script handles:

• Right-click menus — copy user ID / username / message ID / message text / message link
• Reply clicks — smooth-scroll to the original message + 2s highlight flash
• Avatar/name clicks — profile modal with role chips, permissions at close-time, join date
• Reaction clicks — popup with the list of people who reacted
• Theme / language toggles — live switch, no page reload

Data model (excerpt)

interface TicketTranscriptData {
  version: string
  generatedAt: string
  ticket: { id, category?, createdAt?, closedAt }
  guild: { id, name, iconUrl? }
  channel: {
    id, name, topic?, type, parentId?, parentName?,
    viewers?: {
      roles: RoleBrief[],
      members: UserInfo[],     // each with roles[] + permissions[]
      truncated?: boolean
    }
  }
  opener?: UserInfo
  closedBy?: UserInfo
  participants: ParticipantInfo[]
  messages: TranscriptMessage[]   // with attachments, embeds, components,
                                  // stickers, reactions (+ who reacted)
  stats: { totalMessages, userMessages, staffMessages, botMessages,
           attachmentCount, embedCount, reactionCount,
           participantCount, durationMs, durationHuman }
  meta?: Record<string, unknown>
}

Full TypeScript types are shipped in the package (dist/*.d.ts).

Language options

// 1) Show language dropdown for every supported language (default)
languages: 'all'

// 2) Lock to a single language — dropdown hidden
languages: 'ko'

// 3) Offer a curated subset — dropdown visible with only these
languages: ['en', 'ko', 'ja', 'zh-CN', 'zh-TW']

// Plus: which language to show first
defaultLanguage: 'en'

Built-in language codes:

| | | | | |---|---|---|---| | en — English | ko — 한국어 | ja — 日本語 | zh-CN — 简体中文 | | zh-TW — 繁體中文 | es — Español | pt-BR — Português (BR) | fr — Français | | de — Deutsch | it — Italiano | ru — Русский | tr — Türkçe | | pl — Polski | nl — Nederlands | ar — العربية | hi — हिन्दी | | id — Bahasa Indonesia | vi — Tiếng Việt | th — ไทย | uk — Українська | | cs — Čeština | sv — Svenska | fi — Suomi | el — Ελληνικά |

Message text and user identifiers are never translated — only UI chrome, summary cards, modal labels, permissions, context menu, and time units.

Requirements and caveats

• Bot intents — enable GuildMessages, MessageContent, GuildMembers, GuildMessageReactions. Otherwise message content, member data, and reaction-user lookups will be empty.
• Permissions — the bot needs View Channel + Read Message History on the target channel. For reactor lookups, no extra permission is needed beyond reading the channel.
• Asset inlining — inlineAssets: true (default) downloads every attachment; large channels can produce multi-MB HTML files. Cap via maxAssetSize. Setting inlineAssets: false keeps raw Discord URLs which eventually expire (~24h).
• Rate limits — message fetch pages of 100 are sequential; 10k-message channels take roughly a minute.

License

GPL-3.0