Skip to main content
Glama

Feedbucket MCP Server

IMPORTANT

Deprecated: use the official Feedbucket MCP server. Feedbucket now has its own MCP server: see feedbucket.app/features/mcp. This repository was an unofficial, community-built server and is no longer maintained. Thanks to everyone who used it.

A Model Context Protocol (MCP) server that lets AI assistants (Claude Code, Cursor) manage website feedback from Feedbucket.

Beta Software: This project is in beta. Use at your own risk. Please report issues on GitHub.

Prerequisites

Before installing this MCP server, make sure you have:

  1. A Feedbucket account

  2. Feedbucket installed on your website

  3. Protected mode enabled (recommended):

    • Go to Widget Settings

    • Set "Trigger Feedbucket using a query string" to Yes

    • Save settings and copy your secret key

Related MCP server: Fider MCP Server

Quick Start

1. Install from npm

npm install -g feedbucket-mcp

2. Run the setup wizard

# For protected projects:
feedbucket-setup https://your-website.com YOUR_FEEDBUCKET_SECRET

# For public projects:
feedbucket-setup https://your-website.com

The setup script automatically:

  • Extracts the project ID from your website HTML

  • Fetches the private key from Feedbucket API

  • Configures Claude Code and Cursor

3. Restart your IDE

Restart Claude Code or Cursor to load the new MCP server.

4. Start using it

Ask your AI assistant:

  • "Show me unresolved feedback from Feedbucket"

  • "What's the feedback summary for this project?"

  • "Add a comment to feedback #12345 saying we've fixed this"


Alternative: Clone from GitHub

If you prefer to clone the repository:

git clone https://github.com/swiftcomza/feedbucket-mcp.git
cd feedbucket-mcp
npm install && npm run build
npm run setup -- https://your-website.com YOUR_FEEDBUCKET_SECRET

Available Tools

Tool

Description

feedback_list

List and filter feedback with smart summarization

feedback_get

Get full details of a specific feedback item

feedback_stats

Quick project overview (total, resolved, by type, recent activity)

feedback_comment

Add comments to feedback items

feedback_resolve

Mark feedback as resolved

api_status

Check API connection status

feedback_list

List feedback with powerful filtering:

Parameters:
- resolved (boolean): Filter by resolution status
- limit (number): Items to retrieve (default: 10, max: 50)
- offset (number): Skip items for pagination
- summary (boolean): Return condensed format (default: true)
- page_filter (string): Filter by page URL (partial match)
- reporter_filter (string): Filter by reporter name
- feedback_type (string): 'screenshot', 'video', or 'text'
- created_after (string): ISO date filter

feedback_get

Get complete details for a single feedback item:

Parameters:
- feedback_id (number, required): The feedback ID

Returns full text, all comments, attachments, browser/device info, and console logs.

feedback_stats

Get a quick project health overview:

Returns:
- Total feedback count
- Resolved vs unresolved breakdown
- Feedback by type (screenshot/video/text)
- Recent activity (7 days, 30 days)
- Top 5 pages with most feedback

feedback_comment

Add a comment to a feedback item:

Parameters:
- feedback_id (number, required): The feedback ID
- comment (string, required): Comment text
- reporter_name (string): Defaults to "Claude AI Assistant"
- reporter_email (string): Defaults to "claude@anthropic.com"
- resolve (boolean): Also resolve the feedback (default: false)

feedback_resolve

Mark a feedback item as resolved:

Parameters:
- feedback_id (number, required): The feedback ID

Setup Options

# Protected project - provide URL and secret
npm run setup -- https://your-website.com YOUR_SECRET

# Public project - just the URL
npm run setup -- https://your-website.com

Manual Project ID (if auto-detection fails)

# With project ID only (public)
npm run setup -- --project-id tMbUCuQ4FeFJVViMEPlb

# With project ID and API key (protected)
npm run setup -- --project-id tMbUCuQ4FeFJVViMEPlb --api-key YOUR_SECRET

Configure only one IDE

# Claude Code only
npm run setup -- https://your-website.com YOUR_SECRET --claude

# Cursor only
npm run setup -- https://your-website.com YOUR_SECRET --cursor

Extract credentials without configuring

npm run setup -- https://your-website.com YOUR_SECRET --extract

Full help

npm run setup -- --help

Manual Configuration

If you prefer to configure manually:

Claude Code

Add to .mcp.json in your project root:

{
  "mcpServers": {
    "feedbucket": {
      "command": "node",
      "args": ["/path/to/feedbucket-mcp/dist/index.js"],
      "env": {
        "FEEDBUCKET_PROJECT_ID": "your-project-id",
        "FEEDBUCKET_PRIVATE_KEY": "your-private-key",
        "FEEDBUCKET_API_KEY": "your-api-key"
      }
    }
  }
}

Or use the CLI:

claude mcp add feedbucket \
  -e FEEDBUCKET_PROJECT_ID="your-project-id" \
  -e FEEDBUCKET_PRIVATE_KEY="your-private-key" \
  -e FEEDBUCKET_API_KEY="your-api-key" \
  -- node /path/to/feedbucket-mcp/dist/index.js

Cursor

Add to .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "feedbucket": {
      "command": "node",
      "args": ["/path/to/feedbucket-mcp/dist/index.js"],
      "env": {
        "FEEDBUCKET_PROJECT_ID": "your-project-id",
        "FEEDBUCKET_PRIVATE_KEY": "your-private-key",
        "FEEDBUCKET_API_KEY": "your-api-key"
      }
    }
  }
}

Environment Variables

Variable

Required

Description

FEEDBUCKET_PROJECT_ID

Yes

Project ID from the data-feedbucket attribute

FEEDBUCKET_PRIVATE_KEY

Yes

Private key for commenting/resolving (auto-fetched by setup)

FEEDBUCKET_API_KEY

No*

API key for protected projects (feedbucketKey from URL)

*Required if your project uses query string protection.


Development

# Install dependencies
npm install

# Development mode (with hot reload)
npm run dev

# Type checking
npm run typecheck

# Linting
npm run lint

# Build for production
npm run build

Architecture

This MCP server was reverse-engineered from the Feedbucket browser widget to provide:

  • Smart Summarization: Optimizes large datasets for AI consumption

  • Client-side Filtering: Reduces API calls by filtering locally

  • Full TypeScript: Strict mode with comprehensive types

  • Production Ready: Proper error handling and validation

API Endpoints Used

Method

Endpoint

Purpose

GET

/projects/{id}

Fetch project with all feedback

POST

/feedback/{id}/comments

Add comment to feedback

PUT

/feedback/{id}/resolve

Mark feedback as resolved


License

MIT

Available Tools

6 tools
api_statusB

Check Feedbucket API connection status and configuration

Inputs: no named parameters.

feedback_commentC

Add a comment to a specific feedback item to ask for clarification or provide updates

Inputs: feedback_id (required), comment (required), reporter_name, reporter_email, +1 more.

feedback_getA

Get full details of a specific feedback item including all comments, attachments, and session data. IMPORTANT: For screenshot feedback, the response includes a "resource" field with the screenshot URL. You should download this image to a temporary file and view it to see the visual context including the green dot showing where the user clicked. The screenshot provides crucial context for understanding the feedback.

Inputs: feedback_id (required).

feedback_listA

Fetch all feedback items from the Feedbucket project with intelligent filtering for AI consumption. Automatically optimizes data to prevent overwhelming responses. Each feedback item may include a screenshot URL in the "resource" field - use feedback_get to retrieve full details and view the screenshot for visual context.

Inputs: resolved, limit, offset, summary, +4 more.

feedback_resolveC

Mark a feedback item as resolved after actioning it

Inputs: feedback_id (required), reporter_name, reporter_email.

feedback_statsA

Get a quick summary of project feedback statistics - total count, resolved/unresolved breakdown, feedback by type, and recent activity

Inputs: no named parameters.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.7/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose with no overlap: api_status checks connectivity, feedback_comment adds comments, feedback_get retrieves details, feedback_list fetches items, feedback_resolve marks items as resolved, and feedback_stats provides summaries. The tools target different actions on the feedback resource, making misselection unlikely.

    Naming Consistency5/5

    All tools follow a consistent verb_noun pattern with 'feedback_' prefix for feedback-related operations (e.g., feedback_comment, feedback_get) and 'api_' for the status check. This predictable naming scheme enhances readability and agent usability.

    Tool Count5/5

    With 6 tools, the server is well-scoped for managing feedback in a Feedbucket project. Each tool earns its place by covering essential operations like listing, getting, commenting, resolving, and checking stats, plus an API status tool, without being overwhelming or insufficient.

    Completeness5/5

    The tool set provides complete CRUD/lifecycle coverage for feedback management: list (feedback_list), read (feedback_get), update (feedback_comment, feedback_resolve), and delete is implied through resolution. Additional tools for stats and API status enhance functionality, leaving no obvious gaps for agents to handle feedback workflows.

    Maintenance

    ActivityMaintained
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      Connect your AI assistant to Featuriq — the product feedback and roadmap tool for SaaS teams. Browse top feature requests, search feedback with natural language, update statuses, notify users when features ship, and manage your roadmap — all from your AI client. Authenticates via OAuth. No manual API key setup needed.
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Enables AI-assisted user feedback collection and management via a web interface, with MCP integration for automated replies and tool execution.
      1
      5 npm
      MIT