create_group_from_brief
Build a grounded synthetic audience from a population brief.
Minds · Model Context Protocol
Remote MCP server · Live
Build evidence-grounded synthetic audiences, test decisions, and inspect disagreement from the AI assistant where your work already happens.
The research sequence
Minds grounds a population before the first question is asked, then keeps the study available for follow-up instead of collapsing it into one answer.
Describe a market, buyer group, stakeholder set, or research population in plain language.
External statistics and research shape the audience before synthetic respondents are created.
Test concepts, messages, products, positioning, or hypotheses across the complete panel.
Compare segments, follow disagreement, and move structured results into the next decision.
Connect
Use OAuth when your client supports it. For clients that accept custom headers, generate an API key in Minds under Settings → API Keys.
claude mcp add --transport http mindsai https://getminds.ai/mcp \
--header "Authorization: Bearer minds_YOUR_API_KEY"
Open Settings → Connected Apps, add the remote endpoint, and authorize with OAuth.
Open Customize → Connectors, add the remote endpoint, and complete OAuth authorization.
Add https://getminds.ai/mcp as a remote MCP server in the client’s MCP settings.
Research tools
The server describes its current schemas to every connected client. These are the core operations an assistant can orchestrate.
create_group_from_brief
Build a grounded synthetic audience from a population brief.
ask_group
Interview, survey, or poll one saved audience.
create_panel
Create a panel or study around a research goal.
ask_panel
Ask one focused question across the complete panel.
plan_panel_study
Turn a research goal into a structured multi-question plan.
run_panel_study
Execute an approved plan as a durable server-side study.
get_panel_study
Inspect progress, outputs, and method artifacts.
export_panel
Export findings in common research and data formats.
Scope
Minds does not replace representative human research. It improves the first pass: expose objections, compare reactions, and decide which assumptions deserve validation with real respondents.
Open the complete setup guideVersion 2.0.0 is preserved as an open software record: Minds MCP Integration Reference on Zenodo.