create_contact
ChatGPTCreate a contact in the user's workspace, for someone who isn't one yet: search list_people by name first, since a name that is already a contact is almost always that person. Give a name, an email, or both. An email that already belongs to a contact gives that contact instead. Without an email, a name that contacts already have creates nothing and lists them: ask the user whether it is one of them, then pass same_as with that ID, or different_person: true once the user says this is someone else. Returns the contact ID.
create_folder
ChatGPTCreate a new folder. Specify a parent via parent_folder_id; if omitted, the folder is created at the root of the workspace — or in your private folder when the workspace ("team folders") is disabled for the organization. Use list_transcriptions or get_folder_hierarchy to look up the parent folder ID first.
create_glossary
ChatGPTCreate a glossary in the workspace: terms that transcriptions and translations should spell consistently (names, brands, jargon). Apply it to a file with glossary_ids on create_transcription or retranscribe. organization_default applies it to every new file in the workspace, for teammates too.
create_summary_template
ChatGPTCreate a new meeting summary template. The body is markdown where each heading becomes a section in the generated summary (e.g. "## Action Items\n## Key Decisions\n## Discussion Summary"). Optionally include meeting_context to give the AI additional instructions about the type of meeting.
create_transcription
ChatGPTCreates a transcription or subtitles from a publicly accessible media file URL (e.g., a direct link to an audio/video file, a YouTube or Vimeo link, or a public cloud storage share link). The file is imported and processed in the background — check progress and retrieve the result with list_transcriptions or get_transcription. To transcribe local files, direct the user to upload them at https://www.happyscribe.com.
delete_folder
ChatGPTSoft-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.
delete_summary_template
ChatGPTDelete a summary template. Only the template creator or a workspace admin can delete it. The template is soft-deleted and can be recovered within 10 days.
delete_transcriptions
ChatGPTSoft-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.
export_transcription
ChatGPTExport transcripts as downloadable files and return a download link. Formats: txt, md, docx, pdf, srt, vtt, json, csv, xlsx and html, with options for timestamps, speaker names, comments and highlights; srt and vtt are subtitle files. Several transcripts in one export come back as one zip. Links last 24 hours. Translation is not available.
finalize_upload
ChatGPTPart of the HappyScribe upload box, which calls it itself. Attached files go to create_transcription in media_file.
get_company
ChatGPTGet one company from the user's connected HappyScribe workspace: its profile, the people from it the user can see in their own recordings, and the user's recent recordings with that company.
get_export_status
ChatGPTCheck an export that export_transcription reported as still processing, and get its link.
get_folder_hierarchy
ChatGPTGet 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.
get_glossary
ChatGPTGet a glossary with all its terms. Returns the full term list including source text, translations, and context notes.
get_helpdesk_article
ChatGPTFetch one HappyScribe helpdesk article by ID and return its title, full body, and URL. Use this when you already know the article ID. To find an article by keyword, use search_helpdesk first.
get_meeting_diagnostics
ChatGPTGet 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.
get_person
ChatGPTGet detailed profile for a specific person including their company, transcription statistics, and recent transcriptions.
get_summary_template
ChatGPTGet the full details of a summary template by ID (for user/workspace templates) or slug (for system templates). Returns the name, markdown body, sections, meeting context, and visibility.
get_transcription
ChatGPTGet detailed information about a specific transcription
get_transcription_status
ChatGPTCheck the progress of a transcription, by the order_uuid from create_transcription or by a transcription_id. Returns one status (importing, analyzing, transcribing, needs_credits, waiting_for_payment, done or failed) and the next step. While it runs, poll_after_seconds says when to check again. HappyScribe sends no notification in the chat when it finishes: the transcript is ready once this reports done, and the user can ask for a new check anytime.
get_transcriptions
ChatGPTGet detailed information about multiple transcriptions at once. Use this after listing transcriptions to get full content/summaries for multiple files efficiently.
get_upload_url
ChatGPTPart of the HappyScribe upload box, which calls it itself. Attached files go to create_transcription in media_file.
get_video_frames
ChatGPTExtract visual frames (screenshots) from a video recording at specific timestamps. Use this when the conversation references something visual — a screen share, presentation, diagram, or UI — and you need to see what was on screen. Returns the frames as images. Only works for video recordings (meetings, screen recordings); audio-only files have no frames.
get_workspace
ChatGPTGet information about the current workspace this connection has access to
list_calendar_events
ChatGPTList 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. The current year is 2026.
list_companies
ChatGPTBrowse and search companies in the knowledge graph. Returns companies with their people count and transcription statistics.
list_glossaries
ChatGPTList glossaries in the workspace. Glossaries define terminology that should be used consistently in transcriptions and translations — preferred spellings, proper nouns, domain-specific terms, etc.
list_people
ChatGPTBrowse and search the people from recordings in the user's connected HappyScribe workspace. Returns only contacts the user can see, with their company and how many recordings they appear in.
list_read_files
ChatGPTList 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.
list_summary_templates
ChatGPTList meeting summary templates available in the workspace. Returns your private templates, shared workspace templates created by teammates, and built-in system templates. Templates define the structure and sections of AI-generated meeting summaries.
list_transcriptions
ChatGPTList transcriptions accessible to the user with optional filtering. Returns results ordered by creation date (newest first). A good starting point to see recent transcriptions. When the user asks about their own transcriptions (e.g. "my files", "my meetings", "meetings I was in", "calls I had this week"), pass scope: 'mine' — it 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 it, results include every file the user can access via folder permissions. If there are too many results to browse, get_folder_hierarchy shows the folder structure so you can narrow down with folder_path. Supports filtering by folder (folder_path), location, date range, person, company, and pagination. When filtering by person or company, list_people or list_companies can help find the exact name first; person_email is the most reliable filter.
listen_to_audio
ChatGPTListen to a specific segment of a transcription's audio with an audio model and answer a question about it. Use this when you need to hear what actually happened in the audio — tone of voice, background noise, overlapping speakers, pronunciation, music, non-verbal cues, or anything the transcript text cannot capture. Requires start_time and end_time in seconds (max 5 minutes per clip). Use timestamps from the transcription content to target the segment of interest.
move_transcriptions
ChatGPTMove 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.
reassign_speakers
ChatGPTReassign speakers for one or more time ranges in a transcription. The server splits affected paragraphs at word boundaries and assigns the given speaker label to all words in each range. Adjacent paragraphs with the same speaker are merged automatically. Use this to fix diarization errors where paragraphs contain speech from multiple people. Returns the resulting speaker timeline so you can verify the changes.
regenerate_summary
ChatGPTRegenerate the meeting summary for a transcription. Optionally specify a template to use — otherwise the existing template (or the default from the resolution chain) is used. The summary is generated asynchronously; use get_transcription to check the result.
rename_folder
ChatGPTRename a folder.
rename_speakers
ChatGPTRename one or more speakers in a transcription. Pass a mapping of current speaker label -> new speaker label, e.g. { "Speaker 1": "Alice", "Speaker 2": "Bob" }. All paragraphs whose speaker matches a key are updated. Match is exact (case-sensitive). Use get_transcription first to see the current speaker labels.
rename_transcription
ChatGPTRename a single transcription.
replace_text_batch
ChatGPTApply several find-and-replace fixes to one transcription in a single new version. Several fixes in one call make one version, where replace_text_in_transcript makes one per fix. Pairs run in order, each on the result of the ones before (so a later find can match text an earlier replace produced). Matching is exact and case-sensitive. The optional time window applies to every pair. Pairs that match nothing are reported and skipped.
replace_text_in_transcript
ChatGPTFind and replace exact text in a transcription. Replaces every occurrence of find with replace across all paragraphs. Optionally constrain to a time window with from_seconds and to_seconds — only occurrences whose word-level timestamps intersect that window are replaced. Use this to fix consistent typos, names, or product mentions. For speaker label changes, use rename_speakers instead.
request_missing_capability
ChatGPTRecords a request that no available tool can handle, so the HappyScribe team learns what to build next. The capability is described in general terms (for example "translate a transcript into German" or "send the summary to Slack"), without names, emails or other personal details. It is not for requests an available tool covers, nor for errors that already name a next step (a missing summary, a file still processing, no minutes left). One call per request is enough. It does not add new tools.
retranscribe
ChatGPTRe-runs automatic transcription (ASR) on the existing media of a transcription already in the workspace, replacing the current transcript in place (the transcription keeps its ID). This replaces everything: transcript text, speaker diarization, and speaker labels. Any edits the user made — including renamed speakers — will be lost (recoverable from version history inside the editor). There is no way to reprocess only one aspect; retranscription always re-runs the full pipeline. Use this to fix a file that was transcribed in the wrong language, or to re-process it after its settings changed. Only finished automatic transcriptions with their media still available can be retranscribed. Retranscription runs in the background — poll with get_transcription (display_mode: metadata_only) until the state is "done".
search_helpdesk
ChatGPTSearch 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.
search_transcriptions
ChatGPTSearch 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.
set_meeting_template
ChatGPTConfigure which summary template to use for future meetings. With scope "meeting": sets the template on a calendar event (and all upcoming instances if recurring). With scope "default": sets the template as your personal default for all future meetings.
share_transcription
ChatGPTShare a transcription: invite people by email with view or edit rights, and/or turn the public link on (view only or editable) or off. Invited people get an email from HappyScribe, so it fits an explicit request to share, with the exact emails and view/edit rights confirmed by the user. A public link lets anyone with it open the file, so it fits an explicit request for one. Leaving public_link out keeps the link as it is. Returns the link to share.
show_transcript
ChatGPTShow a transcript to the user as a card in the chat, with an audio player and the first paragraphs, which expand in place. It fits a user who wants to see, open, read or listen to a finished transcript, including right after an upload finishes. It works for files that get_transcription_status reports as done. One card shows one transcript, so several files take one call each. get_transcription returns the text for you to read.
show_upload_box
ChatGPTShow an upload box in the chat where the user can drop or pick one or more audio or video files from their device, on desktop or mobile. It fits any request to upload, send or transcribe a file that has not been attached in a form you can pass yourself, including a plain "I want to upload a file", and attachments in clients that give no file access. Documents and images are out of scope. The box uploads each file, starts its transcription and shows progress itself, so its files need no create_transcription call. Once it shows "Uploaded", the user can ask anytime whether the transcript is ready.
unlock_transcription
ChatGPTUnlock a transcription that get_transcription_status reported as needs_credits, spending the workspace credits it costs so it continues. It spends the user's credits, so it needs the user's explicit agreement for this file. If the workspace still cannot afford it, the user can upgrade in their HappyScribe account; adding minutes does not resume the file by itself, it still has to be unlocked.
update_company
ChatGPTEdit a company in the user's workspace. Send only the fields to change; an empty string clears industry, city or country. A new domain leads and the old one still matches its contacts; a domain another company owns is refused. Every contact at the company sees the change. get_company shows the current values.
update_contact
ChatGPTEdit a contact in the user's workspace. Send only the fields to change; an empty string clears a field, the name and email excepted. A contact's email can be added, or changed while no meeting uses it; a workspace member's email can't be changed. get_person shows the current values.
update_glossary
ChatGPTEdit a glossary: rename it, change its languages or default status, and add, edit or remove terms. Terms to edit or remove are matched by their current text (case-insensitive). Use get_glossary first to see the current terms.
update_summary_template
ChatGPTUpdate an existing summary template (name, body, meeting_context, or visibility). Only the template creator can edit it. Omitted fields are left unchanged. The body is markdown where each heading becomes a section in the generated summary. Visibility is "private" (only you) or "workspace" (all team members).
verify_quotes
ChatGPTVerifies quote text against the actual transcription content and returns precise timestamps and working links to each quote in the editor. Recommended when extracting quotes — links or timestamps generated from memory are usually incorrect. Counts as a file read (same quota as get_transcription with full_text).
create_folder
Claudecreate_summary_template
Claudecreate_transcription
Claudedelete_folder
Claudedelete_summary_template
Claudedelete_transcriptions
Claudeget_company
Claudeget_folder_hierarchy
Claudeget_glossary
Claudeget_helpdesk_article
Claudeget_meeting_diagnostics
Claudeget_person
Claudeget_project
Claudeget_summary_template
Claudeget_transcription
Claudeget_transcriptions
Claudeget_video_frames
Claudeget_workspace
Claudelist_calendar_events
Claudelist_companies
Claudelist_glossaries
Claudelist_people
Claudelist_projects
Claudelist_read_files
Claudelist_summary_templates
Claudelist_transcriptions
Claudelisten_to_audio
Claudemove_transcriptions
Claudereassign_speakers
Clauderegenerate_summary
Clauderename_folder
Clauderename_speakers
Clauderename_transcription
Claudereplace_text_in_transcript
Claudesearch_helpdesk
Claudesearch_transcriptions
Claudeset_meeting_template
Claudeupdate_project_notes
Claudeupdate_summary_template
Claudeverify_quotes
Claude