Introduction
Collect complaints from people and AI agents, and manage them in one workspace.
Complaintr turns feedback from almost any channel into structured reports. People write in a widget, a chat app, or a GitHub comment. AI agents call a tool. Your own software sends an HTTP request. Every report lands in one inbox, attached to the application it is about.
How it works
You register an application for each product you want to hear about. Its name is the address every channel uses. Then you connect the channels your users already use: the web widget, an MCP client, Telegram, Slack, GitHub, Instagram, or the HTTP API.
Conversational channels are handled by the Complaintr assistant. It asks what happened and what the person expected, confirms the summary, and files the report. When triage is enabled, each report also gets a category and a severity.
Reports arrive in Complaints, where you filter, read, resolve, and reopen them. The assistant can summarize what is open, and a report can be shared as an image card when you choose.
Concepts
- Application: a product, service, or project that receives feedback. Names are unique across Complaintr.
- Complaint: one report with a title, description, source, status, and optional triage. Feature requests are complaints too.
- Channel: a way in, such as the widget, an MCP client, or a chat app.
- Source: the label that shows which channel or agent filed a complaint.
- Assistant: the AI behind every conversational channel and the Chat page.
- Survey: a set of questions sent through personal links and answered in a conversation.
Start here
Quickstart
From a new account to your first resolved complaint.
Integrations
Connect the channels your users already use.
Complaints
Filter, read, resolve, and reopen reports.
API and tools
HTTP endpoints, MCP tools, and limits.
Beta
Complaintr is under active development. Do not use it for production or critical workflows yet.