Dovetail MCP server
Dovetail in any MCP client, through Memtro: Dovetail does not publish this server. Memtro's Dovetail connector reads Dovetail through its API with the credentials you connect, and one endpoint hands those tools, every other system you connect and Memtro's memory tools to Claude Code, Cursor, Claude Desktop or anything else that speaks Streamable HTTP, as you.
What the model can do
Dovetail customer research: projects, the research data in them (interviews, survey responses, tickets) by title and date with the content of one entry, highlights with their text and tags, tags, docs (research write-ups and reports) with their content, and workspace members. The contacts database of participants is not read. Can add data to a project, write a doc, comment on a doc, create a project or a tag, tag a highlight, and delete data or a doc, each only when asked.
Reads
dovetail_projectsDovetail research projects: title, folder and when created, newest first.dovetail_dataResearch data in Dovetail (interviews, survey responses, support tickets, recordings): title, project and when it was added, newest first by default.dovetail_highlightsHighlights marked in Dovetail research data: the highlighted text (cut at 1000 characters), the tags applied, the data entry it is in…dovetail_data_contentThe content of one Dovetail data entry (an interview transcript, notes, a survey response) as Markdown, cut at max_chars (default 12,000).dovetail_usersDovetail workspace members (staff): name, e-mail, job title, role (contributor, manager, viewer) and whether a workspace admin.dovetail_tagsDovetail tags used to code highlights: title, whether a project or workspace tag, its project, board and group.dovetail_docsDovetail docs (research reports, insights and write-ups): title, project or folder and when created, newest first.dovetail_doc_contentThe content of one Dovetail doc as Markdown, cut at 12,000 characters: the findings, recommendations and evidence a team wrote up.
Acts, when you ask
These change something in Dovetail. The model is told to use them only when you ask for exactly that, and each one is granted by name: per team on a shared connection, by the owner on a personal one.
dovetail_create_dataAdd a data entry (notes, a transcript, a ticket's text) to a Dovetail project, with a title and plain-text or HTML content…dovetail_create_docWrite a new Dovetail doc (a report or write-up) with a title and Markdown, HTML or plain-text content, in a project, in a folder…dovetail_comment_on_docPost a page comment on a Dovetail doc as the key's user, or a reply in an existing thread with thread_id.dovetail_create_projectCreate a Dovetail project, optionally inside a folder and from a workspace template (which brings its fields, views and tags).dovetail_create_tagCreate a Dovetail tag: in a project with project_id, otherwise on a workspace tag board (needs a plan with workspace tags).dovetail_tag_highlightApply up to 10 tags to a Dovetail highlight on a data entry.dovetail_delete_dataDelete a Dovetail data entry, with its highlights.dovetail_delete_docDelete a Dovetail doc.
Connect it
First connect Dovetail on Dashboard → Connections, then add Memtro to your client once. The Dovetail tools appear alongside the tools of everything else you have connected; nothing is configured per connector.
Claude Code
The Memtro plugin registers the server and a skill for when to use it; the tools are then available in every session.
claude plugin marketplace add memtro-org/claude-memtro
claude plugin install memtro@memtro --config api_key=sk-memtro-...Cursor
Add the server to .cursor/mcp.json in the project, or to the global file under ~/.cursor/.
{
"mcpServers": {
"memtro": {
"url": "https://platform.memtro.com/api/mcp",
"headers": {
"Authorization": "Bearer sk-memtro-..."
}
}
}
}Claude Desktop
Create a personal API key on Dashboard → Profile → API keys. In Claude Desktop add an MCP server of type HTTP with URL https://platform.memtro.com/api/mcp and the header Authorization: Bearer sk-memtro-…
OpenCode
A remote server in opencode.json.
{
"mcp": {
"memtro": {
"type": "remote",
"url": "https://platform.memtro.com/api/mcp",
"headers": {
"Authorization": "Bearer sk-memtro-..."
}
}
}
}Pi
One command: pi mcp add memtro --url https://platform.memtro.com/api/mcp --bearer-token-env-var MEMTRO_API_KEY. Tools arrive as mcp__memtro__*. The pi-memtro extension registers it for you, alongside Memtro as the model.
Any other client
Add an MCP server of type HTTP (Streamable HTTP, stateless) at https://platform.memtro.com/api/mcp with the header Authorization: Bearer sk-memtro-.... Codex, Zed, Windsurf, Cline and the SDKs all take a remote server this way.
Who may use it
The key identifies the person, and the server offers that person the tools they may use everywhere else in Memtro. A shared Dovetail connection is used by the teams an administrator granted it to, tool by tool, on the Permissions page; a personal connection is used by its owner alone, and its tools that act stay off until the owner allows them. See Teams and permissions and Connectors.
Ask it like this
- › What have customers said about onboarding?
- › Which interviews were added to the pricing project this month?
- › How often does the slow export tag come up?
- › Summarise the latest research report on churn
- › Add my notes from today's call with BigCo to the pricing project
Questions
- Is this an official Dovetail MCP server?
- No. It is Memtro's MCP server with Memtro's Dovetail connector behind it: the connector reads Dovetail through its API with the credentials you connect on Dashboard → Connections, and Dovetail publishes nothing here. The same endpoint carries every other system you connect and Memtro's memory tools.
- What does it cost?
- Nothing on top of Memtro. One price per organisation by its number of users, from $50 a month for up to 10 users after a 7-day free trial, with every connector, the MCP server, jobs and memory included. You bring your own model keys.
- Can it change anything in Dovetail?
- Only when you ask, and only what it has been granted: add a data entry (notes, a transcript) to a project; write a new doc, in a project, a folder or the workspace; comment on a doc, as the key's user (workspace members see it); create a project; create a tag in a project or on a workspace tag board; apply tags to a highlight; delete a data entry (to the project's trash for 30 days); delete a doc (to the trash for 30 days). Every other tool reads, searches or lists. An administrator grants each acting tool by name per team, and a personal connection acts only after its owner allows it.
- Which clients work?
- Any MCP client that speaks Streamable HTTP: Claude Code, Cursor, Claude Desktop, OpenCode, Codex, Zed, Windsurf, Cline and the Anthropic and OpenAI SDKs. There is no local process to run; the server is remote and authenticated with your key.