Overview

Userback MCP (Model Context Protocol) lets your developer tools and AI agents connect to Userback and automatically load relevant context (feedback details, comments, project/workflow info, console logs, and network logs) directly into your prompt flow.

As a team, you can debug faster without manually copying IDs, recreating long issue timelines, or switching constantly between tools.

Connect the Userback MCP endpoint into supported MCP clients like ChatGPT, Cursor, Claude Desktop, Claude Code, VS Code, Windsurf, and OpenCode, then authenticate to start using Userback context where you work.

Prerequisites

1. Configure Apps & Tools

Claude Desktop

Steps:

1. Open Search and tools.
2. Click + Add connectors.
3. Select Userback from the connector directory.
4. Connect and run auth.
5. Sign in to Userback and approve access for your workspace.

ChatGPT

1. Open the App Store in ChatGPT.
2. Go to Settings.
3. Open Apps & Connectors.
4. Select Userback and click Connect.
5. Sign in to Userback and approve workspace access.

claude mcp add Userback https://mcp.userback.io/v1/mcp/ -t http -s user

Cursor

One-click install:

Add Userback to Cursor

Manual setup:

JSON

{
  "mcpServers": {
    "userback": {
      "url": "https://mcp.userback.io/v1/mcp/"
    }
  }
}

After saving, restart Cursor if needed, then complete OAuth authorization in your browser.

{
  "servers": {
    "Userback": {
      "url": "https://mcp.userback.io/v1/mcp/",
      "type": "http"
    }
  }
}

{
  "mcpServers": {
    "Userback": {
      "url": "https://mcp.userback.io/v1/mcp/"
    }
  }
}

2. Getting Started

Once configured:

1. Start a prompt in your MCP client.

2. Ask your agent to use Userback MCP.

3. Complete OAuth authorization when prompted.

   

4. Approve the MCP tool calls the agent needs.

5. Iterate in focused steps: search -> inspect -> decide -> update.

3. Available Features in Userback MCP

Userback MCP supports:

• Finding and reviewing feedback across projects
• Searching by filters and by intent
• Reading related technical diagnostics (console/network logs)
• Posting comments and updating feedback status/ownership/details
• Organizing and triaging work with project and workflow context

4. Tips

Prompting tips:

1. Start with one clear goal per prompt (for example: "summarize", "prioritize", or "update").
2. Be specific about what you want (project, date range, status, or theme).
3. Review suggested updates before applying them.
4. Keep requests focused and iterative instead of asking for everything at once.

5. Examples

Examples of how teams can use Userback MCP:

Product Engineer uses Userback MCP to debug a bug report with context from feedback details and logs.

Review Userback feedback ID 12345, analyze likely root cause from details + logs, cross-reference with the codebase, and propose an implementation plan.

Product Manager uses Userback MCP to identify patterns across multiple feedback items and produce actionable work packages.

Analyze all "In Progress" feedback in the App project in Userback, cluster recurring themes, estimate impact, and propose efficient implementation batches to minimize context switching.

Product Designer uses Userback MCP to pass design feedback into an agent and generate a concrete implementation plan.

Build an implementation plan for any requested UX changes in the App project and include acceptance criteria for QA.

FAQ