user_feedbackLABSCO SUMMARY
One line in your custom prompt turns “done” into “does this look right?”
Adding an instruction to request feedback before completing a task is what makes this work at all, and it is the difference between an agent that marks a broken build finished and one that shows you the running app first. For multi-step start commands, put them in a task runner and point the saved command at that rather than trying to chain them in one string.
What it is
A small MCP server that opens a feedback window: the agent describes what it did, you look, optionally run the project's own command, and type back what is wrong. Built for desktop-app work, where checking the result means actually using the thing.
What you get
user_feedback— takes a project directory and the agent's summary, and returns your reply- A per-project
.user-feedback.jsonholding the command to run, saved from the UI - An execute-automatically option, so the command runs the moment the window opens instead of waiting for a click
- A human-in-the-loop step that works in Cline and Cursor without either of them supporting one natively
Requirements
uv installed globally, and the repository cloned somewhere your client can point at. Add the server to your client config with a generous timeout — the call is open for as long as you take to answer — and auto-approve the feedback tool so it does not need a confirmation click every time.
Setup effort
Build from source — git clone https://github.com/mrexodia/user-feedback-mcp.git
USE CASES
01
Reviewing a desktop app change before the agent moves on
Let it run the dev command, look at the window, and type the correction straight back into the same task.
02
Cutting the loop where the agent declares victory too early
Add the feedback instruction to your custom prompt so every task ends with a question rather than a claim.
03
Skipping the click on every iteration
Turn on automatic execution in the project config so the app is already running when the window appears.
1 tool
Listed in project docs. Your client may show a different list.
BEFORE YOU INSTALL
What to know first
Tools you get
1 tool named in the project's documentation.
Requires
- uv installed globally
- A clone of the repository — the client runs server.py from it
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
Install uv globally — on Windows run pip install uv; on Linux or macOS run curl -LsSf https://astral.sh/uv/install.sh | sh
TERMINAL
git clone https://github.com/mrexodia/user-feedback-mcp.gitIn Cline, open MCP Servers → Installed → Configure MCP Servers, which opens cline_mcp_settings.json, and add the block from the Cline tab
MCP Clients
Cline
CLINE · JSON
{
"mcpServers": {
"github.com/mrexodia/user-feedback-mcp": {
"command": "uv",
"args": [
"--directory",
"c:\\MCP\\user-feedback-mcp",
"run",
"server.py"
],
"timeout": 600,
"autoApprove": [
"user_feedback"
]
}
}
}c:\MCP\user-feedback-mcp is the README's example location; yours is the full path of your clone. Inside JSON a Windows backslash is written twice, as shown.
