Features

Your feedback data, inside your AI agent.

Install the MCP server and your AI agent can query, triage, and act on feedback without leaving the conversation. Ask "what are users most frustrated with?" — get a cited answer from your actual data.

Works with Claude, Cursor, and more

Compatible with Claude Desktop, Claude Code, Cursor, and any MCP-compatible AI agent. One config file, instant access.

Read and write tools

List, search, and filter feedback. Update status, tags, and assignments. Bulk-update up to 100 items. Scope-aware — write tools only appear with write permissions.

Semantic search built in

Ask natural language questions. The MCP server uses vector embeddings to find relevant feedback by meaning, not just keywords.

Agent workflows

Built-in prompts for triage sessions and weekly reviews. Your agent can review critical feedback, identify duplicates, and summarize trends autonomously.

How it works.

1

Add to your agent config

One JSON block in your Claude Desktop, Cursor, or Claude Code settings. Point it at your API key and project ID.

2

Your agent discovers the tools

The MCP server registers tools automatically. Your agent sees list_feedback, search_feedback, get_stats, and more.

3

Ask questions, take action

"What are the top 5 critical bugs?" "Tag all login issues as auth-bug." Your agent queries and updates feedback natively.

Common questions.

What agents are supported?

Any MCP-compatible agent — Claude Desktop, Claude Code, Cursor, and the growing list of tools implementing the Model Context Protocol.

Can agents modify my data?

Only with a write-scoped API key. Read-only keys only expose query tools. You control the permissions.

Is there a setup guide?

Yes. The docs have step-by-step instructions for Claude Desktop, Cursor, and Claude Code with copy-paste config examples.

More features.

See it in action.

30-day free trial. No credit card. Your first weekly brief arrives Monday.

Start 30-day free trial

No credit card required