Skip to main content
Glama
Minds MCP Server

Minds MCP Server

Official
by minds-ai-co

Minds MCP Server

DOI

Run AI market research from your assistant: build synthetic Audiences and run Studies for concept testing, message testing, and segment comparison.

Endpoint: https://getminds.ai/mcp (streamable HTTP) Docs: https://getminds.ai/mcp/setup Quick-start guide: https://minds-ai-co.github.io/minds-mcp/ Technical reference: https://minds-mcp.readthedocs.io/en/latest/ GitBook: https://minds-1.gitbook.io/minds-mcp/ Client compatibility: https://minds-mcp-compatibility.pages.dev/ Setup generator: https://minds-mcp-setup-generator.netlify.app/ Public API explorer: https://minds-api-explorer.vercel.app/?utm_source=content&utm_medium=content&utm_campaign=content-seo-mcp-reference&utm_content=mcp-repository-api-explorer API + MCP integration lab: https://minds-api-mcp-integration-lab-bba685.gitlab.io/?utm_source=content&utm_medium=content&utm_campaign=content-seo-mcp-reference&utm_content=mcp-repository-integration-lab Archived release: https://doi.org/10.5281/zenodo.21711429 Auth: OAuth 2.1 with PKCE (dynamic client registration or Client ID Metadata Documents), or a Minds API key (Authorization: Bearer minds_…)

This repository is documentation for a hosted MCP server. There is no code to install or run. The server is operated by Minds and the implementation lives in our application, not here.

Documentation maintenance

reference/ is the single source for maintained integration content. Read the Docs and GitBook publish the same Markdown through repository sync. GitHub Pages, Cloudflare Pages, and the Netlify setup generator are intentionally thin, platform-specific surfaces; Zenodo preserves versioned release snapshots. See DOCUMENTATION.md for the publishing contract.

Related MCP server: Sablier MCP Server

What it does

Minds is a synthetic market research platform. This MCP server lets ChatGPT, Claude, Cursor, and any other MCP client run customer research end to end without leaving the assistant.

Describe an audience in a brief ("German Gen Z grocery shoppers", "enterprise IT buyers in fintech") and the server runs deep web research to ground that audience in government statistics, peer-reviewed studies, and industry reports rather than in generic model priors. Then ask the Audience a question, run a structured Study, and export the results.

Typical jobs:

  • Concept testing and message testing before creative production

  • Ad pretesting, landing page audits, and packaging tests with per-segment reactions

  • Buyer persona validation and objection discovery for B2B go to market

  • Segment comparison and positioning checks across markets

  • Screening research hypotheses before commissioning fieldwork with real respondents

Studies run durably server side, so a long study survives the chat session that started it. Results export to PDF, DOCX, PPTX, Markdown, CSV, XLS, and SPSS SAV.

Scope. Minds does not replace representative human fieldwork. It replaces the slow first pass: sharpening the question, surfacing objections, and deciding which assumptions deserve real-respondent validation.

Setup

ChatGPT

Enable developer mode under Settings → Security and login, then open Plugins and select the plus button. Add https://getminds.ai/mcp as a public HTTPS connection, choose OAuth, and sign in to Minds. Availability depends on your account and workspace policy.

Study results can render as interactive widgets inline. See the official OpenAI connection guide and the Minds setup guide.

Claude Desktop and claude.ai

Customize, then Connectors, then add https://getminds.ai/mcp as a remote connector and authorize via OAuth.

Claude Code

claude mcp add --transport http mindsai https://getminds.ai/mcp \
  --header "Authorization: Bearer minds_YOUR_API_KEY"

Cursor

Settings, then MCP, then add https://getminds.ai/mcp and authorize via OAuth. The repository also contains a Marketplace-ready Cursor plugin manifest at .cursor-plugin/plugin.json; its MCP configuration is intentionally kept in cursor.mcp.json so the shared Agent Plugins configuration can retain the portable streamable-http transport name.

Google Antigravity

agy plugins install https://github.com/minds-ai-co/minds-mcp

The root plugin.json and mcp_config.json make this repository an Antigravity plugin. Antigravity discovers OAuth from the hosted endpoint; authenticate under Settings, then Customizations. To add only the server, run agy mcp add --type http minds https://getminds.ai/mcp or put {"mcpServers": {"minds": {"serverUrl": "https://getminds.ai/mcp"}}} in ~/.gemini/config/mcp_config.json. Full guide: https://getminds.ai/guide/integration-antigravity

Gemini CLI remains supported for Google Cloud enterprise and API-key users. Its legacy extension manifest stays available as gemini-extension.json.

ChatGPT and Codex Plugins

The repository includes the shared plugin package at .codex-plugin/plugin.json and its remote server configuration in .mcp.json. The public directory listing uses the same hosted endpoint and maintained reference content.

VS Code (Copilot), Windsurf, Langdock, Open WebUI

Add https://getminds.ai/mcp as a remote MCP server. Full per-client instructions, including API key setup, are at https://getminds.ai/mcp/setup.

API keys

Generate one in the Minds app under Settings → API Keys. Keys start with minds_ and are passed as Authorization: Bearer minds_....

Tools

The server advertises 52 tools. Clients read the current schemas at runtime; the full list with access levels is in reference/tools.md.

Tool

What it does

list_audiences

For a listing or search reply, include both the returned top-level workspaceUrl and each listed Audience's workspaceUrl verbatim in inline code: scheme, query and trailing slash.

list_formations

Lists Audience segmentations for filters; may save defaults and recover builds, replacing subgroup memberships.

import_audience_sources

Imports supplied UTF-8 text, Markdown, CSV and JSON research files into account-owned storage.

create_audience_from_brief

Builds grounded Audience/Mind profiles with exact segment allocation.

get_audience_creation_progress

Read-only; never starts or retries work.

render_audience_review

Shows source research, distributions, proposed Minds, the draft and training for one or more Audience operations, display only: the user confirms or asks for changes in chat.

get_audience_limits

Returns the authenticated account's current plan (planName, as the app names it), its Audience size allowance (per-Audience cap including any team or contract allowance, customSizeMax, per-mode ceilings) and the synthetic responses left, read live; it supersedes limits stated elsewhere.

ask_audience

Accepts one standalone question request for an Audience, restricted to supplied segments, and creates a private Study.

export_audience

Exports an Audience brief or kind validation_report (scores, KPIs, answer shares, provenance) as Markdown/PDF/DOCX/PPTX.

duplicate_audience

Adds an independent copy of an Audience, its Minds and retained knowledge.

recalibrate_audience

Changes an existing Audience in place: it keeps its id, its Minds keep theirs (except copies below), and no Mind is deleted.

manage_audience_support

Manage owner-issued, expiring staff access to one Audience’s composition and receipts.

validate_audience

Starts or cancels a validation of an Audience against real survey results: the best-fitting published surveys it finds, listed surveys named in surveys.benchmarkIds, your own survey files (questionnaire and results, asked as given), or your own aggregated results passed as benchmark.

get_audience_validation

Reads how closely an Audience matches real surveys.

list_studies

List or search saved Studies, or read an authorized team catalog, result feed or tracker snapshot with source identities, full-filter totals and page continuation.

create_study

Creates a separate, new Study workspace when the user requests a new workspace.

ask_study

Requests exactly one respondent-visible question: one standalone question, or one adaptive follow-up whose wording could not be known before earlier results.

get_study_status

Read Study progress, saved replies, uncalibrated forecasts and exports without generating replies.

request_study_answers

Request actual replies (default 8; all=true for all).

cancel_study_run

Stop pending Study work; keep saved replies.

render_study_results

Completed Study results or labeled experimental, uncalibrated forecasts.

export_study

Queues Study results export: executive/full reports in PDF/DOCX/PPTX/Markdown; raw CSV/XLS/SPSS SAV.

duplicate_study

Permanently copy a Study with its questions and results over the same Audiences.

export_heatmap

Exports a completed website heatmap by Study/message ID as a ZIP of stored screenshots and overlays.

get_study_heatmap

Read saved attention by assetKey and, for monadic Studies, variantId.

start_study_heatmap

Start or retry analysis of an assigned original file, image, video or public website, including a monadic variantId.

export_mind

Exports an existing Mind by ID/name: inline Markdown by default, or PDF/DOCX/PPTX.

get_shared_mind_knowledge

Returns stored public Mind sources/assessments by share ID, including shares outside the private workspace.

create_study_plan

Saves a non-executing research draft in an existing Study, including a multi-question concept test or survey.

revise_study_plan

Revises the saved draft at current draft.id and draft.revision without starting research.

get_study_plan

Reads the latest saved revision and full review of one draft.

attach_study_audiences

Attaches existing accessible Audiences without removing others, planning or running research.

render_study_plan

Shows a saved Study draft at its current revision for interactive review and confirmation.

start_study_run

Runs one reviewed draft revision after explicit confirmation in a later user turn; skipping review is not confirmation.

resume_study_run

Resumes an existing Study run after explicit user approval: a plan-limited run after the required paid upgrade, or a failed comparison question via retryQuestionIndex (only its failed groups, with the original sources and respondents; also after the run was stopped, which keeps everything else stopped).

cancel_study_question

Stops one standalone Study question by runId, or studyId and questionId: queued or preparing stops at once, being answered stops at its next step.

update_item

Renames a Study or Audience or moves it into a folder.

set_sharing

Turns a Study's or Audience's public link or team sharing on or off.

list_research_methods

Lists the research-method catalog without changing or running methods.

list_study_drafts

Lists a page of unfinished Study draft summaries, newest first, or returns the complete saved planning state for one exact draft ID.

list_study_templates

Lists owned/team-shared template summaries, most used first; templateId returns full configuration, revision, questions, response settings and attachments.

enterprise_use_cases

Read paged team business use cases and role-scoped counts of attributed drafts, Studies and completed runs.

save_enterprise_use_case

Save a revision-checked team business use case with objectives, template revisions, planned targets and lifecycle state.

save_study_draft

Creates or checkpoints an unfinished Quick or Custom Study draft without starting research.

save_study_template

Creates a private Custom template with its questions, response settings, attachments and method.

update_study_template

Updates one owned Custom template at update.expectedRevision.

use_study_template

Creates an independent editable Study draft from a readable template at use.expectedRevision.

delete_study_template

Only for saved Study templates; it cannot delete Studies, Audiences, Minds or answers.

read_study_summary

Reads saved whole-Study Markdown, evidence, generation state and your refresh receipt.

refresh_study_summary

Queues a whole-Study summary refresh after explicit approval.

plan_study_questions

Deprecated alias of create_study_plan and revise_study_plan; identical behaviour.

run_study_questions

Deprecated alias of start_study_run; identical behaviour.

Server card: https://getminds.ai/.well-known/mcp/server-card.json (minds-ai).

Call list_research_methods for current executable status, requirements, and fallbacks before choosing a research method.

Discovery

  • Server card: https://getminds.ai/.well-known/mcp/server-card.json

  • OAuth protected resource metadata: https://getminds.ai/.well-known/oauth-protected-resource/mcp

  • Registry name: ai.getminds/minds

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers