MCP Server
Use AI assistants to manage your UserHero workspaces, projects, and feedback
MCP Server
The UserHero MCP (Model Context Protocol) server allows AI assistants like Claude, GitHub Copilot, and Cursor to interact with your feedback data programmatically.
What is MCP?
Model Context Protocol is an open standard that lets AI assistants connect to external tools and data sources. With UserHero's MCP server, you can:
- Query feedback - Ask your AI assistant to summarize recent feedback or find specific issues
- Manage projects - Create and configure projects through natural conversation
- Update feedback status - Triage and resolve feedback without leaving your IDE
- Configure widgets - Adjust widget settings through AI commands
Key Features
User-scoped API Keys
One API key gives access to all workspaces you belong to
Full RBAC Support
Permissions follow your workspace role (owner/admin/member/viewer)
Audit Logging
All tool invocations are logged for compliance
Rate Limiting
100 requests per minute per API key
Supported Clients
| Client | Status |
|---|---|
| Claude Desktop | ✅ Supported |
| VS Code (GitHub Copilot) | ✅ Supported |
| Cursor | ✅ Supported |
| Windsurf | ✅ Supported |
Example Usage
Once configured, you can ask your AI assistant things like:
"Show me all bug reports from the last week"
"Create a new widget called 'Exit Survey' with NPS template"
"Mark feedback #abc123 as resolved and add a note saying it was fixed in v2.1"
"What's our feedback volume trend this month?"