AI-Powered Workflow

Let AI fix feedback automatically

Connect SuperFeedback to Cursor via MCP. Your AI reads feedback, locates code, makes fixes, and marks items as resolved. Zero manual lookup.

Install MCP in

Opens Cursor and installs automatically

From feedback to fix in 4 steps

1

Get all feedback

get_feedback_summary

AI retrieves all open comments with element selectors, text content, and page context.

2

Get element details

get_comment_details

Full CSS path, XPath, bounding box, and surrounding HTML for precise code location.

3

AI edits code

Cursor edits files

Using element selectors, AI finds the exact component and applies the requested changes.

4

Mark as done

mark_feedback_resolved

After fixing, AI marks the feedback as resolved. Client sees the update instantly.

Three powerful MCP tools

get_feedback_summary

View all feedback with counts - see what needs attention. Returns comment ID, text, status, element info, and page path.

Returns:
  • Total comment count
  • Open (unresolved) count
  • List of comments with IDs

get_comment_details

Full context for any comment. Get CSS selectors, XPath, bounding box, replies, and design suggestions.

Returns:
  • Full element data (selectors, xpath)
  • Section context
  • All replies/thread history

mark_feedback_resolved

Mark feedback as done after making changes. Updates the comment status and notifies clients.

Requires:
  • Admin code (SUPER_FEEDBACK_ADMIN_CODE)
  • Comment ID to resolve

Get started in 2 minutes

One-Click Install

Install MCP in
After installing:
  1. Open Cursor Settings → MCP
  2. Find super-feedback in the list
  3. Replace your-admin-code-here with your Admin Code
  4. Restart Cursor to apply changes

Manual Configuration

Add this to your Cursor MCP settings:

mcp.json
{
  "super-feedback": {
    "command": "npx",
    "args": [
      "-y",
      "super-feedback-mcp@latest"
    ],
    "env": {
      "SUPER_FEEDBACK_ADMIN_CODE": "your-admin-code-here"
    }
  }
}
Where to find your Admin Code: Go to your SuperFeedback project settings. The Admin Code is required for the mark_feedback_resolved tool to work.

Why use MCP integration?

10x Faster Fixes

No more switching between dashboard and code. AI reads and acts on feedback directly.

Perfect Context

Element selectors, CSS paths, and XPath give AI everything it needs to locate code precisely.

Auto-Resolution

Mark items as resolved programmatically. Clients see updates immediately.

Batch Processing

Ask AI to fix all open feedback in one go. Handle routine items in minutes.

Ready to automate feedback fixes?

Install the MCP server and let AI handle the repetitive work.