Feedbk

You can run and use an MCP server that provides AI-moderated interviews and surveys directly in your workflow. This server exposes an HTTP endpoint you connect to with your MCP client, enabling you to design, deploy, and collect insights from interviews and surveys powered by AI.

How to use

You interact with the Feedbk MCP server by connecting your MCP client to the remote HTTP endpoint and then issuing natural language requests to design interviews or surveys. The AI interviewer automatically asks intelligent follow-up questions and helps you refine questions to gather deeper insights. You can create interview guides, export them, and share interview links with participants for remote participation.

How to install

Prerequisites: you need an MCP client compatible with HTTP MCP connections. The following steps show how to add the Feedbk MCP server to your client and start using it.

claude mcp add --transport http feedbk https://mcp.feedbk.ai/mcp

If you use a code editor or IDE with MCP integration, you can also configure Feedbk as a remote server in your environment. Examples include VS Code, Claude Desktop, OpenAI ChatGPT, Google Gemini CLI, and OpenAI Codex CLI.

"mcp": {
  "servers": {
    "feedbk": {
      "type": "http",
      "url": "https://mcp.feedbk.ai/mcp"
    }
  }
}

Additional setup and usage notes

Use the remote Feedbk MCP server to create interview guides by describing what you need. The system will guide you through naming the interview, stating its purpose, and composing the questions.

Practical examples

Create a customer satisfaction survey about your mobile app by asking your MCP client to design it. Or build a user research interview to understand why users cancel subscriptions. Or design an employee feedback survey about remote work preferences.

Available tools

AI Probing

Automatic intelligent follow-up questions based on participant responses to deepen insights without manual configuration.

Single-Choice Question

Question type where respondents select one option from a list.

Multiple-Choice Question

Question type where respondents select multiple options from a list.

Open-Ended Question

Question type that accepts free-form textual responses.

Rating (Likert Scale)

Question type that captures ratings on a defined scale from 1 to 5 or similar.

Interview Design

Capability to create and tailor interview or survey names, purposes, and question sets.