home / mcp / user prompt mcp server

User Prompt MCP Server

Provides a terminal-driven user input workflow for AI agents via a pop-up prompt with structured Q&A responses.

Installation
Add the following to your MCP client configuration file.

Configuration

View docs
{
  "mcpServers": {
    "jacob-j-thomas-user-context-retrieval-mcp-server": {
      "command": "C:\\Tools\\UserPromptMcpServer\\UserPromptMcpServer.exe",
      "args": []
    }
  }
}

You can run a dedicated MCP server that lets AI agents ask you clarifying questions mid-task through a pop-up terminal window. This server exposes a single tool called user_prompt, which opens a new window on your machine to collect user answers and return them to the agent for continued execution.

How to use

To use this MCP server, configure your MCP client to point at the local or remote server, then let the agent call the user_prompt tool during its task. When invoked, you see a separate terminal window that presents a clear reason for the user input and a numbered list of questions. You answer each question in order, and the agent receives the responses to continue its task with the additional context. If you close the window or do not respond within ten minutes, the agent gets a fallback message and can decide how to proceed.

How to install

Choose one of the three installation paths that fits your setup and follow the steps end-to-end.

Option A: GitHub Releases
1. Download the appropriate zip:
   - Windows (x64): UserPromptMcpServer-win-x64.zip
   - macOS (Intel): UserPromptMcpServer-osx-x64.zip
   - macOS (Apple Silicon): UserPromptMcpServer-osx-arm64.zip
   - Linux (x64): UserPromptMcpServer-linux-x64.zip
2. Extract to a permanent location, for example:
   - Windows: C:\Tools\UserPromptMcpServer\
   - macOS/Linux: ~/tools/UserPromptMcpServer/
3. Configure your MCP client (example for Claude Code CLI):
   claude mcp add user-prompt -- "C:\Tools\UserPromptMcpServer\UserPromptMcpServer.exe"
4. Updating: download the new release and replace the files in the same folder.

Option B: .NET Global Tool
1. Install the tool (requires .NET 8.0 SDK or later):
   dotnet tool install -g UserPrompt
2. Configure your MCP client (Claude Code CLI example):
   claude mcp add user-prompt -- user-prompt
3. Updating:
   dotnet tool update -g UserPrompt

Option C: Build from Source
1. Requires the .NET 8.0 SDK or later.
2. Clone and publish:
   git clone https://github.com/Jacob-J-Thomas/user-context-retrieval-mcp-server.git
   cd user-context-retrieval-mcp-server
   dotnet publish UserPromptMcpServer -c Release -r win-x64 -o ./publish
   Replace win-x64 with your platform, e.g., osx-arm64 or linux-x64.
3. Configure your client the same way as Option A, pointing at the executable in ./publish/.

Additional setup notes

The server uses stdio transport. The MCP client launches the server process and communicates over stdin and stdout. Each installation option yields a usable command you can configure in your client’s MCP settings.

Tool usage and behavior

You will use a tool named user_prompt. When invoked, the agent supplies a reason and a list of questions. You respond to each question in sequence. The terminal window shows a header, the agent’s reason, the numbered questions, and prompts like 1>, 2>, etc. After you provide all answers, the window may close automatically, and the agent will process the results.

Notes on terminal behavior

If you do not answer within ten minutes or you close the prompt window, the agent receives a clear fallback notice and can choose how to proceed.

Available tools

user_prompt

Requests user input for a specific reason and a list of questions, returns answers as clearly formatted Q&A pairs for the agent to continue execution.