MCP App Store
by KeenEthics
Dovetail icon

Dovetail MCP Server for Claude and ChatGPT

By Dovetail

Overview

Connect your customer intelligence to Claude and put real customer data behind every answer.

How to connect

  1. Claude: Open the Dovetail listing in the Claude connectors directory and choose Connect. Claude listing Dovetail exposes 9 tools here.
  2. ChatGPT: Open the Dovetail listing in the ChatGPT app directory and choose Connect. ChatGPT listing Dovetail exposes 63 tools here.

Tools

add_project_context_doc

ChatGPT
Link an insight doc to a project's AI context. Requires project_id and doc_id. Max 10 docs.

add_project_context_keywords

ChatGPT
Add keywords to a project's AI context. Requires project_id and keywords. Duplicates are ignored. A project can have at most 50 keywords.

add_workspace_context_doc

ChatGPT
Link an insight doc to the workspace AI context. Requires doc_id and a workspace admin. Max 10 docs. Duplicates are ignored.

apply_tags

ChatGPT
Apply one or more existing tags (up to 10 per call) to an existing highlight. Tags are labels used to categorise highlights during qualitative analysis — for example "Usability Issue" or "Feature Request". Requires highlight_id (resolve via get_project_highlights, get_highlight, or search_workspace) and tag_ids (resolve via list_tags, or create one first with create_tag). Each tag must belong to the highlight's project or to a workspace tag board linked to that project — if any tag ID is invalid or from another project, the whole call is rejected and no tags are applied. Only highlights on data entries can be tagged: a range carried by a doc/insight, or an inline comment anchor, is rejected and will never become taggable — do not retry it. Applying a tag the highlight already has is a safe no-op, so retrying a call is fine. Returns the updated highlight including its full tag list.

complete_file_upload

ChatGPT
Finalize a file upload started with upload_file, after the file's bytes have been PUT to the presigned upload_url. The server verifies the bytes in cloud storage and marks the file as uploaded. Once completed, pass the file_id to import_file_to_data or import_file_to_doc to add the file to a project as data or a doc. The file's processing status may still be pending afterwards — poll get_file to check. Safe to retry: completing an already-finalized file returns its metadata. Requires data:file_import or doc:file_import.

create_channel_datum

ChatGPT
Send a new data point to a Dovetail channel for automated AI analysis. Each data point represents one piece of customer feedback — an app review, support ticket, NPS response, product review, or churn reason — that the channel classifies against its topics and themes. Requires the target channel_id (discover via list_channels or search_workspace), the feedback text, and an ISO 8601 timestamp recording when the feedback was originally received. Optionally provide source_title and source_url for provenance. After creation, the point is queued for asynchronous classification; inspect results via list_channel_data, get_channel_datum, or list_channel_themes.

create_comment

ChatGPT
Post a comment on a doc. Comments are discussion threads on docs — feedback, questions, and follow-ups. Requires the doc_id (resolve via list_docs, get_doc, or search_workspace) and the comment body. Defaults to plain text; pass body_type: "html" for rich-text content. If the doc has no comment thread yet, one is created automatically. Pass optional thread_id from list_doc_comments or get_doc_comment to reply on that page-level or inline thread. The comment is attributed to the authenticated user. Returns the created comment including its ID. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash). Use this when the user wants to leave or add a comment on a doc.

create_data

ChatGPT
Create a new research data entry within a Dovetail project. Data entries hold raw qualitative material — interview transcripts, survey responses, support tickets, CRM records, and similar primary research — and are the surface where highlights and tags are applied during analysis. Requires a project_id; resolve a project from its title or URL first via get_dovetail_projects or search_workspace. Optionally provide an initial title and content body (plain text or HTML), and set custom fields via the fields array. Returns metadata for the new entry; use get_data_content to read its body or list_project_data to enumerate siblings.

create_doc

ChatGPT
Create a new doc in your Dovetail workspace. Docs are rich-text deliverables used to write up research findings, share reports, and publish summaries. Provide the initial body via the content field as HTML, Markdown, or plain text (set content_type accordingly — Markdown is ideal for porting content from Notion, Confluence, or other wikis). Place the doc inside a project (project_id) or a folder (folder_id), but not both; omit both to create it at the workspace root. Returns the new doc's metadata; use get_doc_content to read the body afterwards.

create_folder

ChatGPT
Create a new folder in your Dovetail workspace. Folders organise projects, docs, channels, dashboards, agents, and other folders into a hierarchical structure, like a file system. Provide a title; optionally pass parent_folder_id to nest it inside an existing folder (resolve folder IDs via list_folders or get_folder), or omit it to create the folder at the workspace root. Returns the new folder including its ID.

create_project

ChatGPT
Create a new project in your Dovetail workspace. Projects are containers that organize qualitative research — each one holds data entries, docs, highlights, tags, and insights for a specific research initiative (e.g. an onboarding study or churn investigation). Optionally pass a template_id (discover available templates via list_project_templates) to pre-populate the project with that template's fields, views, tags, and structural layout. Optionally pass a folder_id to place the project inside a specific folder; otherwise the project is created at the workspace root. Returns the new project object including its ID, which can then be used as the parent for create_data, create_doc, and similar tools.

create_tag

ChatGPT
Create a new tag. Tags are labels used to categorise highlights during qualitative analysis — for example "Usability Issue", "Feature Request", or "Positive Feedback". You must pass project_id and/or tag_board_id so the tag lands where the user intends: pass project_id (resolve via get_dovetail_projects or search_workspace) to create a project tag in that project — the same label in two projects is two separate tags. Pass a workspace board's tag_board_id (find IDs via list_tag_boards) to create a workspace tag — one shared label usable across every project linked to that board. Only create a workspace tag when the user has explicitly asked for one: it mutates taxonomy shared across the whole workspace. The tag lands in the first group of the resolved board unless group_id narrows it further. Once created, apply the tag to highlights when calling create_transcript_highlight (via tag_ids) or in the Dovetail web app. Returns the new tag including its ID.

create_transcript_highlight

ChatGPT
Create a highlight on an audio or video transcript by marking a start/end time range. Highlights are passages flagged as significant during qualitative analysis and can be tagged for categorisation. Requires note_id (the data entry that owns the transcript — discover via list_project_data, get_project_data, or search_workspace) plus start_time and end_time in seconds. Optionally pass tag_ids (from list_tags) to apply tags on creation. Bounds are snapped to the nearest word, so timestamps returned by get_highlight or get_project_highlights round-trip cleanly. To pick precise word-level bounds, inspect the data entry's HTML export — each transcript word is wrapped in a span with data-monologue-start-time / data-monologue-end-time attributes. Returns the created highlight including its ID and tags.

download_file

ChatGPT
Get a short-lived presigned URL to download the raw content of a file attachment by its unique identifier. Files are uploaded assets (images, documents, audio, video) embedded in docs, data entries, or insights. Returns a presigned cloud-storage URL (valid for 1 hour) that can be fetched directly with no additional authentication, plus its expiry timestamp. Use get_file first to discover a file's metadata and processing status; use this tool when you need the actual file bytes. Call again to mint a fresh URL if the previous one has expired.

edit_doc

ChatGPT
Update a Dovetail doc's metadata: its title or its folder. This does not change the doc's body text — use edit_doc_content for that. Only the fields you pass are changed. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash).

edit_doc_content

ChatGPT
Edit the body of a Dovetail doc: insert_blocks (markdown at start/end or before/after a block), replace_block (markdown), delete_block and format_text. First call get_doc_blocks; pass its version as expected_version and its block_refs in operations. Refs belong to the read they came from, so put every change in one batch. Operations apply in order: don't target a block an earlier one replaced or deleted — put its formatting in the replacement markdown. The batch is atomic: if any operation fails, nothing changes. Only editable blocks can change; Dovetail modules never do. Mention a member in the markdown with @[<email>] (from list_users); @[<user id>] or @[<full name>] also work. Resolving needs the user:read scope. Only the first 100 tokens in the request are rewritten; later tokens stay written as @[<ref>]. A resolved mention notifies that member, and an unresolved one stays plain text @<ref>. A block that holds a mention is no longer editable, so write its final wording in the same batch. In a list, reword an item by replacing its p block, delete an item by deleting its li, and add one with insert_blocks after an li using '- text'. Blocks that don't fit where they would go (e.g. a table inside a table cell) are refused rather than restructuring the blocks around them. The title isn't part of the body; use edit_doc. Replacing or deleting a block with comments or highlights is refused, and this tool cannot override that: edit around those blocks, or ask the user to make that change in Dovetail. After a success or a 409, call get_doc_blocks again before editing more. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash).

get_channel

ChatGPT
Retrieve detailed information about a specific channel by its unique ID, including its topics. Topics are sub-categories that the channel uses to automatically classify incoming customer feedback. Use this tool after discovering a channel via list_channels or search_workspace to understand how the channel is organized and what topics it tracks. The response includes the channel's title, creation date, folder, and a complete list of topics with their titles and descriptions. Dovetail URL: matches /channels/<id> — extract the channel ID from the URL path.

get_channel_datum

ChatGPT
Retrieve a single channel data point by its ID, including its full text content, sentiment, summary, source URL, metadata, and the themes it has been classified under. Use this after discovering a data point ID via list_channel_data or search_workspace to inspect a specific piece of customer feedback in detail. Dovetail URL: matches /channels/<channel_id>/data/<id> and the themed variant /channels/<channel_id>/ideas/<theme_id>/data/<id> (or the pre-rollout /channels/<channel_id>/themes/<theme_id>/data/<id>) — extract the datum ID (the final segment after /data/).

get_contact

ChatGPT
Retrieve a single contact by its unique identifier. Contacts represent people in the contacts database — interview participants, customers, or research subjects — and may carry custom field data (e.g. company name, Salesforce ID). Use this when you have a contact ID referenced from another resource and need to resolve it to a name, email, or specific custom field value. Dovetail URL: matches /contacts/<id> — extract the contact ID from the path segment after /contacts/.

get_data_content

ChatGPT
Export and retrieve the complete content of a research data entry (such as interview transcripts, survey responses, usability test notes, or field study observations) in markdown format. This tool provides access to the raw or processed content of primary research data, making it readable and analyzable. Use this when you need to review the actual content of customer feedback, analyze specific responses or quotes, or extract detailed information from research sessions. The markdown format ensures the content is well-structured and easy to process for analysis.

get_doc

ChatGPT
Retrieve detailed metadata for a single doc by its unique ID, including title, custom fields, project association, folder, attached files, and cover image. The response does not include the doc's text content — use get_doc_content to read the rich-text body. Use this when you already know a doc ID and need its full metadata before deciding whether to fetch the content. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash).

get_doc_blocks

ChatGPT
Get a Dovetail doc's body as an ordered list of blocks, plus the doc's version. Call this before edit_doc_content and pass the version as expected_version. Each block has a block_ref, type, depth, heading level, plain text (no formatting or links, cut at 2,000 characters and then ending in …) and editable flag. Don't rewrite a block from cut-off text with replace_block: read it in full with get_doc_content first. A container's (list, quote, table, toggle) text repeats the text of the blocks inside it, one line per block. editable=false marks Dovetail modules (labelled in module) and blocks containing or inside one: you cannot change them, but you can insert before or after a module (not inside it). Refs starting with module: are labels only; anchor on the block next to them. If has_more is true, call again with offset=next_offset. If you received fewer blocks than returned_blocks, call again with offset set to this response's offset plus the blocks you received, and a smaller max_chars. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash).

get_doc_comment

ChatGPT
Retrieve a single comment on a doc by its unique identifier. Returns the comment's body, author, timestamps, and thread status. Use this when you already know both the doc ID and the comment ID — for example to inspect a specific comment referenced from another tool's output. Dovetail URL: matches /docs/<slug>-<id>#c=<comment_id> — extract the doc ID from the path (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash) and the comment ID from the c= hash fragment.

get_doc_content

ChatGPT
Export and retrieve the complete content of a Dovetail doc in markdown format. Docs are rich-text research deliverables; this tool returns the full document body including headings, lists, embedded references, and other formatting. Use this whenever a user asks you to read, summarise, quote, or analyse a Dovetail doc by URL or ID. Pair with list_docs or get_doc to discover the right doc first. The response includes a content_pagination field; if the content you received is shorter than its returned_chars value or ends mid-sentence, your client truncated it — call again with offset set to next_offset (and optionally max_chars) to page through the rest. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash). Prefer this tool over get_doc when the user asks to read, summarise, or quote the doc's contents.

get_dovetail_projects

ChatGPT
Retrieve a comprehensive list of all research projects in your Dovetail workspace to understand the full scope of customer research being conducted. Projects organize related research activities, data collection efforts, and analysis work around specific products, features, or research questions. This tool provides project metadata including names, descriptions, creation dates, team members, and status information. Use this as a starting point to discover available research, identify relevant projects for your analysis, or get an overview of organizational research activities. Essential for workspace navigation and project discovery workflows.

get_field

ChatGPT
Retrieve a single custom field definition by its unique identifier, including its label, type (TEXT, NUMBER, SINGLE_SELECT, etc.), options, rank, and the project it belongs to. Use this after discovering a field ID via list_fields when you need the full definition — for example to interpret field values returned on data or doc entries.

get_file

ChatGPT
Retrieve metadata for a single file attachment by its unique identifier. Files are uploaded assets (images, documents, audio, video) embedded in docs, data entries, or insights. Returns the file's name, MIME type, size in bytes, processing status (pending, completed, failed), author, and creation date. Use this to check the processing status of an uploaded file, or to inspect a file referenced from a doc or data entry. To download the file's content, use download_file.

get_folder

ChatGPT
Retrieve a single folder by its unique identifier, including its title, parent folder, creation date, and a list of immediate child folder IDs. Use this to inspect a folder's position in the workspace hierarchy or to discover its direct subfolders before listing the contents of one of them. Dovetail URL: matches /folders/<id> and its section variants /folders/<id>/projects, /folders/<id>/docs, /folders/<id>/channels, /folders/<id>/dashboards, /folders/<id>/agents — extract the folder ID from the path segment immediately after /folders/.

get_folder_contents

ChatGPT
List all items contained directly within a specific folder — projects, docs, channels, dashboards, agents, and child folders. Use this to browse a folder's contents in a file-explorer-style navigation flow. Each item includes its type, title, creation date, and author. Results are paginated and sortable by creation date or title. Dovetail URL: matches /folders/<id> and its section variants /folders/<id>/projects, /folders/<id>/docs, /folders/<id>/channels, /folders/<id>/dashboards, /folders/<id>/agents — extract the folder ID from the path segment immediately after /folders/. Prefer this tool over get_folder when the user wants to know what's inside the folder rather than the folder's own metadata.

get_highlight

ChatGPT
Retrieve a single highlight by its unique identifier, including its text content, associated data entry, tags, and (for transcript highlights) start/end timestamps. Highlights are selected passages of text or time ranges in audio/video transcripts that have been marked as significant during qualitative analysis. Use this when you already know a highlight ID — for example one returned by get_project_highlights or search_workspace — and need its full detail.

get_insight_content

ChatGPT
Export and retrieve the complete content of a research insight in markdown format for easy reading, analysis, and integration into reports or presentations. This tool returns the full text content of an insight including its structure, findings, recommendations, and formatting. Ideal for when you need to analyze the actual written content of insights, extract specific information, or prepare insights for external sharing. The markdown format preserves formatting and makes content easily readable and processable.

get_ip_allowlist

ChatGPT
Retrieve the workspace's IP allowlist mode (disabled, report_only or enforcing) and its user list, the CIDR entries that govern every browser session. Pass the user list's id to get_ip_allowlist_list for its full metadata. Workspace admins only, on workspaces enrolled in the IP allowlist beta.

get_ip_allowlist_list

ChatGPT
Retrieve one IP allowlist by id: its kind (the workspace's user list, or an API key's dedicated list), name, CIDR entries with labels, and who changed it when. Get list ids from get_ip_allowlist (the user list) or list_ip_allowlist_keys (each key's list). Workspace admins only, on workspaces enrolled in the IP allowlist beta.

get_my_ip

ChatGPT
Show the IP address Dovetail's IP allowlist sees for this request, and its family (v4 or v6). Use this before editing an allowlist, to check whether the caller's own address would still be covered. When the address can't be determined, ip is null and reason says why. Workspace admins only, on workspaces enrolled in the IP allowlist beta.

get_project

ChatGPT
Retrieve a single Dovetail project by its unique identifier, including title, author, creation date, folder location, and its overview in Markdown format. Projects are containers that organize qualitative research — they hold data entries, docs, highlights, tags, and insights. Use this after discovering a project via get_dovetail_projects or search_workspace when you need its metadata and overview. The response includes a content_pagination field; if the overview you received is shorter than its returned_chars value or ends mid-sentence, your client truncated it — call again with offset set to next_offset (and optionally max_chars) to page through the rest. Dovetail URL: matches /projects/<id> — extract the project ID from the path segment after /projects/.

get_project_context

ChatGPT
Get a project's AI context: keywords and linked insight docs. Requires project_id from get_dovetail_projects or search_workspace. Each doc is { id, type: "doc" } only — resolve a title via get_doc, list_docs, or search_workspace. Dovetail URL: matches /projects/<id> — extract the project ID from the path segment after /projects/.

get_project_data

ChatGPT
Retrieve detailed metadata and information about a specific research data entry by its unique ID. Research data entries represent individual pieces of customer feedback such as interview sessions, survey responses, usability test results, or observation notes. This tool provides comprehensive details including participant information, data type, collection date, tags, status, and associated metadata. Use this when you have a specific data ID and need to understand its context, source, or characteristics before analyzing its content.

get_project_highlights

ChatGPT
Retrieve customer feedback highlights and key quotes from a specific research project. Highlights are the most important findings, quotes, and observations extracted from customer interviews, surveys, usability tests, and other research data. Use this tool when you need to understand what customers said about specific topics, gather supporting evidence for insights, or analyze sentiment and themes within a project. Each highlight includes the original source, tags, and context. Supports pagination for large datasets.

get_project_insight

ChatGPT
Retrieve detailed metadata and information about a specific insight document by its unique ID. Insights are structured research findings that synthesize patterns, themes, and conclusions from customer data. They typically include recommendations, supporting evidence, and impact assessments. This tool provides comprehensive details including metadata, status, and associated data sources. Use this when you have a specific insight ID and need to understand its context, source, or characteristics before analyzing its content.

get_tag

ChatGPT
Retrieve a single tag by its unique identifier, including its title, scope (project or workspace), the project it belongs to, its workspace tag board (for workspace tags), and its creation date. Use this when you already know a tag ID — for example one referenced from a highlight returned by get_project_highlights — and need to resolve it to a human-readable label.

get_user

ChatGPT
Retrieve a single workspace member's profile by their unique identifier, including name, email, job title, role (CONTRIBUTOR, MANAGER, or VIEWER), and whether they are a workspace admin. Use this to resolve user IDs returned by other tools — for example to attribute a doc, insight, or highlight to its author.

get_workspace_context

ChatGPT
Get the workspace AI context: linked insight docs. Each doc is { id, type: "doc" } only — resolve a title via get_doc, list_docs, or search_workspace.

import_file_to_data

ChatGPT
Create a new data entry in a Dovetail project by importing a file — interview recordings, PDF/DOCX documents, survey exports, and similar research material. Provide the file in exactly one of two ways: url (a publicly accessible URL that Dovetail downloads itself — pass mime_type if the URL has no file extension) or file_id (a file already uploaded via upload_file + complete_file_upload). Audio and video files (MP3, WAV, M4A, MP4, MOV) are automatically queued for transcription, with the transcript added to the data body once processing completes. Returns the new entry's metadata; use get_data_content to read its body once processed.

import_file_to_doc

ChatGPT
Create a new doc by importing a file — PDF/DOCX documents, audio/video recordings, and similar material. Provide the file in exactly one of two ways: url (a publicly accessible URL that Dovetail downloads itself — pass mime_type if the URL has no file extension) or file_id (a file already uploaded via upload_file + complete_file_upload). Audio and video files (MP3, WAV, M4A, MP4, MOV) are automatically queued for transcription, with the transcript added to the doc body once processing completes. Place the doc inside a project (project_id) or a folder (folder_id), but not both; omit both to create it at the workspace root. Optionally set cover_image_file_id to an uploaded image file. Returns the new doc's metadata; use get_doc_content to read the body once processed.

list_channel_data

ChatGPT
List the raw data points (e.g. app reviews, NPS responses, support tickets) that flow into a specific channel. Use this after identifying a channel via list_channels or get_channel to inspect the underlying customer feedback the channel is analyzing. Each data point includes its source timestamp, text content, sentiment, summary, source URL, and the themes it has been classified under. Pass source_timestamp_from and/or source_timestamp_to (ISO 8601 datetimes, inclusive bounds) to scope results to a time window — e.g. set both to query a fixed range, or one to bound an open-ended range. Results are paginated and sorted by source_timestamp descending by default. Dovetail URL: matches /channels/<channel_id>/data — extract the channel ID from the URL path.

list_channel_themes

ChatGPT
Retrieve all themes for a channel. Themes are AI-generated clusters of data points that share a common subject within a topic. Each theme has a title (a short label summarising the cluster), an optional summary (a longer AI-generated description of what data points in the theme share in common), and a datum_count (the number of data points currently classified into the theme). This tool provides metadata about the themes in the channel, including their title, summary, and datum count. Dovetail URL: matches /channels/<channel_id>/themes — extract the channel ID from the URL path.

list_channels

ChatGPT
Browse and discover all channels in your Dovetail workspace. Channels are automated analysis pipelines that process high-volume customer feedback such as app reviews, NPS responses, support tickets, product reviews, and churn reasons into structured insights. Each channel entry includes metadata like title, creation date, and folder location. Pass folder_id to scope results to channels inside a specific folder, or omit it to list channels across the whole workspace. Use this tool to find channels before drilling into a specific channel's details and topics with get_channel. Supports pagination for workspaces with many channels. Dovetail URL: matches /channels.

list_contacts

ChatGPT
Browse the contacts database in your Dovetail workspace. Contacts are people involved in research — typically interview participants, customers, or survey respondents — and may carry custom field data such as company, role, or external identifiers (e.g. a Salesforce ID). Use this to discover contact IDs for cross-referencing participants across data entries, look up a participant by name, or list all known participants. Pass name to narrow results to contacts whose name contains the given substring (case-insensitive), or omit it to list everyone. Results are paginated.

list_doc_comments

ChatGPT
Retrieve all comments on a specific doc, including inline comments, returned in chronological order. Set include_text_range to true to also receive the selected text for inline comments. Use this after locating a doc via list_docs, get_doc, or search_workspace to read its conversation history. Only published comments are included; results are paginated. Dovetail URL: matches /docs/<slug>-<id> (and /docs/<id> when the doc has no title slug) — extract the doc ID (the base62 or UUID portion after the last dash in the slug, or the full segment when there is no dash). Use this when the user references a doc and is asking about its comments, conversation, or discussion thread.

list_docs

ChatGPT
Browse and discover all docs in your Dovetail workspace. Docs are rich-text research deliverables — reports, write-ups, and shareable findings — that synthesize evidence from data entries, highlights, and other sources. The response contains metadata only (titles, folders, creation dates); use get_doc_content to read the actual document body. Pass project_id to scope to a single project, or omit it to list docs across the whole workspace.

list_fields

ChatGPT
Retrieve all custom fields defined on a project. Fields are typed metadata slots attached to data entries and docs — for example a 'Company Name' text field on customer interviews, or an 'NPS Score' number field. Use this to discover what fields are available before reading or filtering on field values. Requires project_id and field_set_type ('data' or 'doc'); results are paginated and sorted by rank by default.

list_folders

ChatGPT
Browse all folders in your Dovetail workspace. Folders organize projects, docs, channels, dashboards, agents, and other folders into a hierarchical structure similar to a file system. Pass parent_folder_id to enumerate children of a specific folder (note: listing only root-level folders via this filter is not available through MCP — use get_folder_contents on a known root folder, or omit the filter to list folders across the workspace). Pass title to narrow results to folders whose title contains the given substring (case-insensitive). Results are paginated.

list_ip_allowlist_keys

ChatGPT
List every API key in the workspace with its owner, creation and expiry dates, and the IP allowlist that governs it: the shared user list, or the key's own dedicated list. Metadata only; key secrets are never returned. Pass a list id to get_ip_allowlist_list for its entries. Workspace admins only, on workspaces enrolled in the IP allowlist beta.

list_personal_docs

ChatGPT
Retrieve all docs authored by or assigned to a specific user. Use this when you have a user ID (e.g. from get_user or list_users) and want to enumerate the docs they own. The response is metadata only — use get_doc_content to read each doc's body. Supports the same filtering, sorting, and pagination as list_docs.

list_project_data

ChatGPT
Browse and discover all research data entries within a specific project to understand the scope and variety of customer feedback collected. This includes interviews, surveys, usability tests, field studies, and other primary research data. The tool returns metadata for each data entry including titles, types, collection dates, participant details, and status information. Essential for understanding what customer data is available in a project, planning analysis workflows, or identifying specific data sources to investigate further. Supports pagination for projects with extensive research data.

list_project_insights

ChatGPT
Discover and browse all research insights within a specific project to understand the key findings and conclusions drawn from customer research. This tool provides an overview of all insights including their titles, descriptions, status, creation dates, and authors. Use this to get an understanding of what research conclusions have been documented, identify relevant insights for further investigation, or assess the scope of analysis completed within a project. Essential for project overviews and insight discovery workflows. Supports pagination for projects with many insights.

list_project_templates

ChatGPT
List all project templates available in your Dovetail workspace. Templates are reusable blueprints that define a project's fields, views, tags, and structural layout. Use this to discover template IDs that can pre-populate a new project with a standard research setup. Results are paginated.

list_tag_boards

ChatGPT
List tag boards in your Dovetail workspace. A tag board holds the tags and tag groups used to categorise highlights. With no arguments, the workspace's shared tag boards are returned — these are the boards whose tags can be applied across every linked project, and the boards you target with create_tag's tag_board_id. Pass project_id (resolve via get_dovetail_projects or search_workspace) to list a single project's own tag board instead. Each board includes its ID, title, tag count, and tag groups (each with its ID, title, and tag count) in display order. Results are paginated.

list_tags

ChatGPT
Browse and discover tags in your Dovetail workspace. Tags are labels applied to highlights during qualitative analysis — for example "Usability Issue", "Feature Request", or "Positive Feedback" — and come in two scopes, reported by each tag's scope field: project tags belong to a single project, while workspace tags live on a workspace tag board and can be applied to highlights in every project linked to that board. Use this to enumerate a categorisation scheme (which themes researchers track), or to resolve a tag's name to its ID before referencing it elsewhere (e.g. create_transcript_highlight's tag_ids). With no arguments, every tag in the workspace is returned, including workspace tags. Narrow with the project_id filter (only tags belonging to that project) or the tag_board_id filter (only tags on that board — a project board with project_id, a workspace board without). Workspace tags include their workspace_tag_board (id and title). Results are paginated.

list_users

ChatGPT
List all members of your Dovetail workspace, including their name, email, job title, role, and admin status. Use this to discover user IDs — for example to enumerate docs authored by a specific user via list_personal_docs — or to display a directory of teammates. Pass name and/or email to narrow results to users whose name or email contains the given substring (case-insensitive); omit both to list everyone. Results are paginated.

search_workspace

ChatGPT
Perform powerful text-based search across all content types in your Dovetail workspace to quickly find relevant customer feedback, research findings, and insights. This universal search tool can locate specific quotes, topics, themes, or concepts within highlights from customer interviews, research insights and conclusions, data entries like survey responses, organized content channels, tagged content, and thematic analysis results. Use this tool when you need to find all mentions of a specific topic, product feature, user behavior, or research question across your entire research repository. Ideal for cross-project analysis, trend identification, and comprehensive research discovery. You can filter results by content type and use pagination to explore large result sets.

upload_file

ChatGPT
Initiate an upload of a local file's bytes to Dovetail, for files up to 5 GB. This is step 1 of 3 for adding a file to a project as data or a doc: (1) call this tool with the file's name, MIME type, and exact size in bytes; (2) HTTP PUT the raw file bytes to the returned upload_url — the PUT needs no authentication but must include the header x-amz-server-side-encryption: aws:kms and a Content-Type header exactly matching the declared type, or it fails with 403; (3) call complete_file_upload with the returned file_id, then pass the file_id to import_file_to_data or import_file_to_doc. The upload_url expires at expires_at, and abandoned uploads are cleaned up after 24 hours, so complete the flow promptly. If the file is already publicly downloadable, skip this entirely and pass its URL straight to import_file_to_data or import_file_to_doc. Requires data:file_import or doc:file_import; importing the file into data or a doc still requires the matching import scope.

get_data_content

Claude

get_dovetail_projects

Claude

get_insight_content

Claude

get_project_data

Claude

get_project_highlights

Claude

get_project_insight

Claude

list_project_data

Claude

list_project_insights

Claude

search_workspace

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/dovetail"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/dovetail)