Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@OriginalVoices MCP Server ask UK fitness enthusiasts about their preferred workout routines"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
This repository has moved to ovlabs/originalvoices-js. It's archived and no longer updated.
@originalvoices/mcp-server is now published from the originalvoices-js monorepo, alongside the other OriginalVoices JavaScript packages. Issues and pull requests go there.
Package: npm
Docs: docs.originalvoices.ai
Ask questions to a previously saved audience by its ID. Use list_projects first to discover projects and their audiences.
Inputs: audienceId (required), questions (required).
Ask questions to a filtered subset of a project's survey panel. Filters respondents by demographics and survey answers using natural language. Use list_projects first to see available projects.
Inputs: projectId (required), questions (required), filter.
Ask questions to a specific audience using Original Voices AI twins. Use this for ad-hoc audience queries with a free-text audience description.
Inputs: audience (required), questions (required).
List all research projects in the user's organization. Each project contains survey questions and pre-defined audiences. Use this to discover projects before using ask_project or ask_audience.
Inputs: no named parameters.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 4 tools
Each tool targets a distinct action: listing projects, asking saved audiences, asking filtered panels, and asking with AI twins. No overlapping purposes.
All tools follow a consistent verb_noun naming pattern in snake_case, with three using 'ask' for queries and one using 'list' for listing, which is logical.
Four tools is well-scoped for this server's purpose. Each tool adds clear value without redundancy.
The tool set covers listing projects and three distinct query methods (saved audience, filtered panel, AI twins). Minor gaps like audience creation or project management are likely out of scope.
Ask Digital Twins of real people about anything: open and choice questions, any audience.
Test ideas, messages and ads with Human Digital Twins built from real people, in your AI assistant.
Synthetic audiences built from real people. Ask anyone anything before you decide.
AI-native market research: research studies, persona audience studies, analysis, reports and media