Cursor Feedback Loop

v1.0.0Developer Toolsstable

Save your Cursor monthly quota! Unlimited AI interactions in one conversation via MCP feedback loop.

ai-assistantclaudecursorcursor-extensionfeedback
Share:
28
Stars
0
Downloads
0
Weekly
0/5

What is Cursor Feedback Loop?

Cursor Feedback Loop is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to save your cursor monthly quota! unlimited ai interactions in one conversation via mcp feedback loop.

Save your Cursor monthly quota! Unlimited AI interactions in one conversation via MCP feedback loop.

This server falls under the Developer Tools category on MCPgee, the world's largest MCP server directory with 33,000+ servers.

Features

  • Save your Cursor monthly quota! Unlimited AI interactions in

Use Cases

Extend Cursor quota
Continuous interaction
Feedback loops
jianger666

Maintainer

LicenseMIT
Languagejavascript
Versionv1.0.0
UpdatedMay 19, 2026
Statushealthy
Maintenanceactive

Works with

ClaudeOpenAIwindowsmacoslinux

Installation

Manual Installation

npx cursor-feedback-extension

Configuration

Configuration Details

Config File

claude_desktop_config.json

Performance

Response Metrics

Response Time< 200ms
ThroughputMedium

Resource Usage

Memory UsageLow
CPU UsageLow

How to Set Up and Use Cursor Feedback Loop

Cursor Feedback Extension is an MCP server and companion VS Code/Cursor extension that creates a structured human-in-the-loop feedback channel within AI agent sessions. The AI agent invokes the interactive_feedback MCP tool to pause and present a summary of its work, the user provides text feedback, images, or file references through a sidebar panel, and the agent continues — all within a single conversation without burning additional Cursor fast-request quota. Developers use it to guide long-running AI tasks interactively and efficiently without restarting sessions.

Prerequisites

  • Node.js 16 or later
  • Cursor IDE or VS Code with MCP support
  • The cursor-feedback VS Code extension installed (ID: jianger666.cursor-feedback)
  • npm or npx available on your PATH
1

Install the Cursor extension

Install the companion extension from the Cursor/VS Code marketplace. Search for 'cursor-feedback' by jianger666 or install it via the command line.

cursor --install-extension jianger666.cursor-feedback
2

Install the MCP server globally (optional)

For a persistent installation, install the cursor-feedback npm package globally. Alternatively, use npx in the config to avoid a global install.

npm install -g cursor-feedback
3

Add the MCP server to Cursor's configuration

Edit the Cursor MCP configuration file at ~/.cursor/mcp.json to register the feedback server. Using npx ensures you always run the latest version.

4

Restart Cursor and verify the server is running

Restart Cursor after saving the config. Open the MCP panel in settings to confirm the cursor-feedback server shows as connected. The interactive_feedback tool should appear in the tool list.

5

Instruct the AI to use the feedback tool

Tell the AI agent to call interactive_feedback at key checkpoints — for example, before making irreversible changes or when it needs clarification. You can include this instruction in your Cursor rules or system prompt.

Cursor Feedback Loop Examples

Client configuration

Cursor MCP configuration (~/.cursor/mcp.json) to run the feedback server via npx.

{
  "mcpServers": {
    "cursor-feedback": {
      "command": "npx",
      "args": ["-y", "cursor-feedback@latest"],
      "env": {
        "MCP_FEEDBACK_TIMEOUT": "300",
        "MCP_AUTO_RETRY": "true"
      }
    }
  }
}

Prompts to try

System prompt additions and test prompts to exercise the feedback loop.

- Add to your Cursor rules: "Before making any destructive change, use the interactive_feedback tool to summarise your plan and wait for my approval."
- "Refactor the auth module, but pause and ask me before deleting any files"
- "Implement the new API endpoint and use interactive_feedback to show me the changes before writing them"
- "Review my PR diff and use interactive_feedback to present your findings and wait for my input"
- "I want to rename this variable across 30 files — show me a preview first"

Troubleshooting Cursor Feedback Loop

The interactive_feedback tool is not visible in Cursor's tool list

Ensure the cursor-feedback server entry is saved correctly in ~/.cursor/mcp.json and that Cursor has been fully restarted. Open Cursor Settings > Features > MCP to check the server status and see any connection errors.

Feedback window does not appear after the AI calls the tool

The companion extension (jianger666.cursor-feedback) must be installed and active for the sidebar panel to appear. Confirm the extension is enabled in the Extensions panel. The extension polls the MCP server's HTTP API and renders the UI in a sidebar view.

Session times out before the user submits feedback

Increase the MCP_FEEDBACK_TIMEOUT environment variable (default 300 seconds) in the MCP server config, and ensure MCP_AUTO_RETRY is set to true so the server automatically retries timed-out requests.

Frequently Asked Questions about Cursor Feedback Loop

What is Cursor Feedback Loop?

Cursor Feedback Loop is a Model Context Protocol (MCP) server that save your cursor monthly quota! unlimited ai interactions in one conversation via mcp feedback loop. It connects AI assistants to external tools and data sources through a standardized interface.

How do I install Cursor Feedback Loop?

Follow the installation instructions on the Cursor Feedback Loop GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.

Which AI clients work with Cursor Feedback Loop?

Cursor Feedback Loop works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.

Is Cursor Feedback Loop free to use?

Yes, Cursor Feedback Loop is open source and available under the MIT license. You can use it freely in both personal and commercial projects.

Browse More Developer Tools MCP Servers

Explore all developer tools servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.

Quick Config Preview

{ "mcpServers": { "cursor-feedback-extension": { "command": "npx", "args": ["-y", "cursor-feedback-extension"] } } }

Add this to your claude_desktop_config.json or .cursor/mcp.json

Read the full setup guide →

Ready to use Cursor Feedback Loop?

Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.

33,000+ ServersFree & Open SourceStep-by-Step Guides