HappyScribe

Search transcripts, summarize meetings, extract action items, and transcribe new files directly from chat — across ChatGPT, Claude Desktop, Cursor, and more.

От сообществаРаботаетБез входаГлобальныйБесплатноТолько чтение

Что умеет

  • List Transcriptions: List transcriptions accessible to the user with optional filtering. Returns results ordered by creation date (newest first). START HERE to see recent transcriptions. **When the us
  • Search Transcriptions: Search for exact text/keywords within transcription content (like grep). Use this to find specific names, product names, or exact phrases. For browsing by topic, date, or catego
  • Get Transcription: Get detailed information about a specific transcription

Какие данные видит

Нужен ли аккаунт

Не нужен: сервер работает без входа

Search transcripts, summarize meetings, extract action items, and transcribe new files directly from chat — across ChatGPT, Claude Desktop, Cursor, and more.

Как подключить

Как подключить HappyScribe к Claude

Адрес сервера, который нужно вставить в настройках Claude:

https://happyscribe.run.tools

  1. Откройте Claude (сайт claude.ai или приложение для компьютера).
  2. Перейдите в Настройки → Коннекторы (в английском интерфейсе: Settings → Connectors).
  3. Нажмите Добавить пользовательский коннектор (Add custom connector).
  4. Вставьте адрес сервера, скопированный ниже, в поле URL удалённого MCP-сервера и нажмите Добавить.
  5. В чате нажмите + → Коннекторы и включите новый коннектор.

Пользовательские коннекторы доступны на тарифах Free, Pro, Max, Team и Enterprise. На Free можно добавить только один. В Team и Enterprise коннектор сначала добавляет владелец организации в Настройки организации → Коннекторы.

Авторизация

После нажатия Добавить откроется окно входа в сервис. Войдите под своим аккаунтом и подтвердите доступ. Пароль Claude не видит.

Как подключить HappyScribe к ChatGPT

Адрес сервера, который нужно вставить в настройках ChatGPT:

https://happyscribe.run.tools

  1. Откройте ChatGPT в браузере (chatgpt.com). Нужен тариф Plus, Pro, Business, Enterprise или Edu.
  2. Один раз включите режим разработчика: Настройки → Приложения → Дополнительные настройки → Режим разработчика (Settings → Apps → Advanced settings → Developer mode).
  3. Откройте Настройки → Коннекторы и нажмите Создать (Create).
  4. Заполните форму: Название (любое), Описание (одна фраза, что делает сервис), URL MCP-сервера (скопируйте ниже).
  5. В поле Аутентификация выберите OAuth, если сервис требует вход, иначе Нет. Нажмите Создать.
  6. В новом чате откройте + → Приложения/Коннекторы и включите коннектор.

Название раздела у OpenAI менялось (Connectors → Apps/Plugins). Если пункт называется иначе, ищите «режим разработчика» в настройках. В рабочих пространствах Business/Enterprise администратор должен разрешить пользовательские коннекторы.

Авторизация

При первом использовании ChatGPT откроет окно входа в сервис. Войдите и подтвердите доступ.

Как подключить HappyScribe к Cursor

Адрес сервера, который нужно вставить в настройках Cursor:

https://happyscribe.run.tools

Быстрый способ: нажмите кнопку Открыть в Cursor ниже, Cursor сам предложит добавить сервер.

Вручную:

  1. В Cursor откройте Settings → Cursor Settings → MCP и нажмите Add new global MCP server.
  2. Вставьте JSON, скопированный ниже, в открывшийся файл ~/.cursor/mcp.json (для одного проекта: .cursor/mcp.json в корне проекта).
  3. Сохраните файл. Сервер появится в списке MCP; при необходимости пройдите авторизацию по кнопке рядом с ним.

Авторизация

Если сервис требует вход, рядом с сервером в списке MCP появится кнопка авторизации.

Список инструментов сервера (30)

Технические названия из tools/list. Нужны только разработчикам.

list_transcriptionsList transcriptions accessible to the user with optional filtering. Returns results ordered by creation date (newest first). START HERE to see recent transcriptions. **When the user asks about THEIR OWN transcriptions** (e.g. "my files", "my meetings", "what have I been working on", "meetings I was in", "calls I had this week") ALWAYS pass scope: 'mine'. This restricts results to files the user owns, participated in (as speaker or calendar attendee), or that were directly shared with them — mirroring the dashboard "My files" view. Without scope: 'mine', this tool returns every file the user can technically access via folder permissions, which is usually NOT what "my files" means. If there are too many results to browse, use get_folder_hierarchy to see the folder structure, then call this again with folder_path to narrow down to specific folders. Supports filtering by folder (folder_path), location, date range, and pagination. When filtering by person or company, use list_people or list_companies first to find the exact name, then use person_email (preferred) or the verified person_name from the results.
search_transcriptionsSearch for exact text/keywords within transcription content (like grep). Use this to find specific names, product names, or exact phrases. For browsing by topic, date, or category, use get_folder_hierarchy + list_transcriptions instead.
get_transcriptionGet detailed information about a specific transcription
get_transcriptionsGet detailed information about multiple transcriptions at once. Use this after listing transcriptions to get full content/summaries for multiple files efficiently.
list_read_filesList files already read this month and show remaining quota. Some plans have a monthly limit on how many unique files can be read with display_mode: "full_text" — once a file has been read, re-reading it is always free. Summaries and metadata are also always free.
verify_quotesREQUIRED for quote extraction: Verifies quote text against the actual transcription content and returns precise timestamps and working links to each quote in the editor. This is the ONLY reliable way to get accurate quote positions - never generate links or timestamps from memory as they will be incorrect. Always call this tool immediately after identifying quotes to extract. Counts as a file read (same quota as get_transcription with full_text).
list_projectsList projects in the workspace. Projects group transcriptions, instructions, and AI conversations around a specific goal (e.g., a research study, client engagement, story).
get_projectGet details of a specific project: name, instructions, notes, files, and members. Use list_projects first to find the project ID.
update_project_notesUpdate the AI memory for a project. Use this to persist key findings, decisions, and patterns discovered across conversations. Keep notes concise, structured, and factual. Only use when you discover something important that should persist across conversations.
list_peopleBrowse and search people who appear in transcriptions. Returns people from the knowledge graph with their company and transcription statistics.
get_personGet detailed profile for a specific person including their company, transcription statistics, and recent transcriptions.
list_companiesBrowse and search companies in the knowledge graph. Returns companies with their people count and transcription statistics.
get_companyGet detailed profile for a specific company including all known people there, transcription statistics, and recent transcriptions involving this company.
get_workspaceGet information about the current workspace this connection has access to
get_folder_hierarchyGet the folder hierarchy with transcription counts. Shows accessible folders organized by location (Team, Private, Shared with me). USE THIS when list_transcriptions returns more results than you can effectively browse and you need to understand how transcriptions are organized. Then use list_transcriptions with folder_path to browse specific folders.
upload_fileOpens the HappyScribe file uploader widget for the user to upload their audio/video files. Call this tool when the user wants to transcribe their own files (mp3, mp4, wav, etc.). This tool renders a widget - DO NOT just explain what to do, actually call the tool to show the uploader. \n\nAfter calling this tool, explain to the user:\n1. Files uploaded to ChatGPT/Claude are NOT accessible to HappyScribe\n2. They need to use the HappyScribe uploader widget that just appeared above\n3. This is a secure, direct upload to HappyScribe servers\n4. ONLY if the widget doesn't appear, tell them to use the fallback_widget_url link\n5. In the ChatGPT mobile app, file picking doesn't work - direct them to use the fallback_widget_url link to upload in their browser\n\nAfter upload completes, the widget will send you a message with a client_upload_uuid. Use this UUID with the create_transcription tool to start transcription.
create_transcriptionCreates a new transcription from a media file and displays a progress monitoring widget. \n\nTWO WAYS TO USE THIS TOOL:\n1. WITH A PUBLIC URL: If the user provides a URL to a publicly accessible audio/video file (e.g., https://example.com/video.mp4, YouTube link, Dropbox share link), use this tool directly with the url parameter.\n2. WITH AN UPLOADED FILE: If the user has uploaded a file to ChatGPT/Claude, or the file is not publicly accessible, use the upload_file tool first to get a client_upload_uuid, then use this tool with the client_upload_uuid parameter.\n\nThis tool renders a progress monitoring widget that updates in real-time. ONLY if the widget doesn't appear in the chat interface, tell the user to use the fallback_widget_url link to monitor progress in their browser. The widget will send you a message with the transcription uuid when complete. Use that UUID with the get_transcription tool to retrieve the transcription text.
get_meeting_diagnosticsGet detailed diagnostic information about why a meeting was or wasn't recorded. Use this to debug failed recordings, understand bot behavior, and troubleshoot calendar integration issues. Provides information about: calendar event details, notetaker meeting status, bot status, recording configuration, and reasons why the bot didn't join or recording failed.
list_calendar_eventsList calendar events (scheduled meetings) with their recording status. Use this to see what meetings are scheduled, which will be recorded, and prepare for upcoming meetings. Supports date filtering to find meetings in specific time ranges (past, present, or future). Can filter by series_ids to get all instances of recurring meeting series.IMPORTANT: be careful with the dates you use. Some LLMs tend to put the wrong year. This year is 2026.
search_helpdeskSearch HappyScribe helpdesk articles to answer questions about product features, policies, and how-to guides. Use this when the user asks about: how HappyScribe works, product documentation, feature explanations, pricing details, data policies (e.g., "how long are files stored?", "is my data used for training?"), account settings, troubleshooting, or any question that might be answered in help documentation. This searches published help center articles and returns relevant excerpts with links to the full articles.
move_transcriptionsMove one or more transcriptions to a different folder in the same organization. Accepts up to 50 transcription IDs per call. Use list_transcriptions or get_folder_hierarchy to look up the destination folder ID first. Cross-organization moves are not supported. Authorization is checked per transcription and on the destination folder; if any check fails, no transcriptions are moved.
rename_transcriptionRename a single transcription.
delete_transcriptionsSoft-delete one or more transcriptions (move to trash, restorable by the user). Accepts up to 50 transcription IDs per call. Authorization is checked per transcription; if any fails, no transcriptions are deleted. Permanent deletion is not exposed via MCP.
create_folderCreate a new folder. Specify a parent via parent_folder_id; if omitted, the folder is created at the root of the workspace. Use list_transcriptions or get_folder_hierarchy to look up the parent folder ID first.
rename_folderRename a folder.
delete_folderSoft-delete a folder. The folder must be empty (no kept files or subfolders). To delete a non-empty folder, first move or delete its contents.
list_glossariesList glossaries in the workspace. Glossaries define terminology that should be used consistently in transcriptions and translations — preferred spellings, proper nouns, domain-specific terms, etc.
get_glossaryGet a glossary with all its terms. Returns the full term list including source text, translations, and context notes.
list_conversationsList past AI conversations from HappyScribe's in-product chat (Ask AI). Returns conversations ordered by creation date (newest first). Use include_analysis to see AI-generated summaries of what each conversation was about.
get_conversationRetrieve messages from a past HappyScribe in-product AI conversation (Ask AI). Use list_conversations first to find the conversation ID. By default shows user/assistant dialogue with tool call names (but not results). Use include_tool_results to see full tool outputs, include_attachments to see file info. Use offset/limit to page through long conversations.