M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
Q

Quackback

Community
Open-source feedback platform. AI agents can search, triage, comment, vote, and manage changelogs.
How to install ↓ Compare GitHub source repository ↗
Category
Collaboration #48 of 55
Stars
★ 312 Popular
Transport
Streamable HTTP
Runtime
Hosted (no local runtime)
Credentials
No credential needed
License
AGPL-3.0
Last commit
Tools
15
39FMRS · D

The Quackback MCP server is a feature-rich toolset for teams with a Quackback instance. It provides 15 tools covering comprehensive management of feedback posts, comments, changelogs, and roadmaps. While official status is not explicitly stated, it is part of the Quackback project and is open-source licensed. Installation requires a self-hosted or cloud instance, and configuration is straightforward. Risks mainly involve data access permissions and potential misoperations, but overall it is powerful, especially for teams looking to automate feedback workflows with AI agents.

Strongest · Maintenance 10/20 Weakest · Security and permissions 4/20

Reliability
8/20
Security and permissions
4/20
Maintenance
10/20
Documentation
8/20
Setup experience
9/20
Why each score
Reliability 8/20
The manifest lists 15 tools with a coherent schema, but the README claims a '23-tool MCP server', an unresolved contradiction. No CI or committed tests were supplied to verify the init handshake or real behavior, so the happy path is plausible but poorly evidenced; score kept below the 10 anchor.
Security and permissions 4/20
Red-line concern: delete_comment is described as 'hard-delete' with no confirmation parameter or explicit guard. No auth model, least-privilege scoping, data boundaries, or external network behavior is disclosed. Because an irreversible delete is exposed by default without confirmation, security must stay in 0-4 range.
Maintenance 10/20
The repo is not archived, has an AGPL-3.0 license and 58 open issues; the README shows CI and 'PRs welcome', indicating active maintenance. However, no commit history, release cadence, dependency-update or security-response evidence was provided, so a mid-range score is appropriate.
Documentation 8/20
The README clearly covers product installation, Docker and local development, but MCP-specific docs are limited to tool names and one-line descriptions in the manifest. Parameter schemas, auth, limitations and troubleshooting are absent, and the 23-vs-15 tool count contradiction weakens trust.
Setup experience 9/20
Self-hosting steps via Docker, Railway and local dev are present and require PostgreSQL/Redis plus .env. No mainstream-client MCP connection example is provided, and no execution evidence (CI/tests) was supplied, limiting the score below the static maximum.

Static review · not runListed 2026-08-07

Read the FMRS scoring method →

Fit and risk

What it can accessUses the networkChanges third-party account data

Best for

  • Teams using Quackback (self-hosted or cloud) who want AI agents to automate feedback management
  • Developers integrating feedback data with other tools like Jira or Slack
  • Customer support teams needing AI-assisted duplicate detection and theme summarization

Not for

  • Users without a deployed Quackback instance
  • Scenarios requiring non-HTTP transport (e.g., stdio or SSE)
  • Those seeking a full replacement rather than an integration with existing tools

Required permissions

  • Read feedback posts, changelog entries, and comments
  • Create and edit feedback posts, changelogs, and comments
  • Manage post status, tags, owner, roadmap assignment, and merging
  • Add/remove comment reactions and votes

Risks and side effects

  • Sensitive customer feedback may be exposed if the instance uses weak authentication or is publicly accessible
  • Modifying post statuses or content may affect workflows, so use caution
  • Merging operations are irreversible (though unmerge_post exists, it may affect related data)
  • Self-hosted instances require maintenance and updates, otherwise security vulnerabilities may arise

Setup

Before you start

Runtime：Hosted (no local runtime)

DATABASE_URL requiredsecret Required in .env for self-hosted Quackback; the PostgreSQL connection string pointing to your own database.
REDIS_URL required Required in .env for self-hosted Quackback; the Redis-compatible store URL used by BullMQ background jobs.

After configuring your Quackback instance, set the MCP server URL to https://{host}/api/mcp, where {host} is your Quackback instance's hostname (e.g., feedback.yourcompany.com). In your MCP client (e.g., Claude Desktop), add this URL to the mcpServers configuration.

claude_desktop_config.json 
{
  "mcpServers": {
    "quackback": {
      "url": "https://feedback.yourcompany.com/api/mcp"
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "quackback": {
      "url": "https://feedback.yourcompany.com/api/mcp"
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add --transport http quackback https://feedback.yourcompany.com/api/mcp

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

After connecting the client to https://feedback.yourcompany.com/api/mcp, check that tools like search and get_details appear in the tool list, then run a search to confirm it returns feedback posts.

Troubleshooting

  1. Check that the `host` variable correctly points to your Quackback instance
  2. Ensure the instance is publicly accessible or network-reachable
  3. Verify API authentication credentials are valid; the MCP server may require the same authentication as the instance
  4. Check server logs for error messages and ensure the MCP server is properly deployed and running

Things to try

Once connected, you can ask your AI assistant things like:

  • Search all feedback posts with open status
  • Triage this post by setting its status to triaged and adding an enhancement tag
  • Add a comment on this feedback post replying to the user
  • Create a changelog entry for this release and link the related posts

Tools 15

search read-only
Search feedback posts or changelog entries with filtering and pagination.
get_details read-only
Get full details for any entity by TypeID (post, changelog).
triage_post writes
Update a post: set status, tags, and/or owner.
vote_post writes
Toggle vote on a feedback post.
add_comment writes
Post a comment on a feedback post, with threaded reply support.
create_post writes
Submit new feedback on a board.
create_changelog writes
Create a changelog entry (draft or published).
update_changelog writes
Update title, content, publish state, and linked posts on a changelog.
Show 7 more tools
delete_changelog destructive
Soft-delete a changelog entry.
update_comment writes
Edit a comment's content.
delete_comment destructive
Hard-delete a comment and its replies.
react_to_comment writes
Add or remove an emoji reaction on a comment.
manage_roadmap_post writes
Add or remove a post from a roadmap.
merge_post writes
Merge a duplicate post into a canonical post.
unmerge_post writes
Restore a merged post to independent state.

Use cases

AI agents automatically triage and organize customer feedback
Extract feedback from chat or email and create posts
Automatically update roadmaps and changelogs based on feedback
Manage duplicate feedback to keep the list clean
Reply to user comments on feedback posts with real-time updates

Supported clients

Claude Desktop

Listed from the project's documentation, not tested by this site.

Overview

Quackback is an open-source customer feedback collection and management platform, positioned as an alternative to Canny, UserVoice, and Productboard. This MCP server allows AI agents to directly interact with a Quackback instance, performing actions such as searching feedback, triaging posts, voting, commenting, and creating/managing changelogs. It supports self-hosting and offers 24 integrations (Slack, Linear, Jira, etc.).

Similar servers

Compare with other MCP servers

Source revision 8d93fc2fe1c3 Data synced 2026-10-12 Read the FMRS scoring method