Connect FeatQ to Lovable

Lovable is an AI app builder, not an MCP client, so it does not connect to the FeatQ server directly. (MCP, the Model Context Protocol, is the standard coding agents like Claude Code use to call FeatQ tools.) With Lovable the loop is simpler and still tight: install the feedback widget with one prompt, collect votes, then paste the top spec from your board back into Lovable to build it.

Setup: one prompt

Paste this prompt into Lovable. Replace YOUR_BOARD_SLUG with the slug from your board URL.

Add this script tag just before the closing body tag of my app: <script src="https://featq.com/widget.js" data-board="YOUR_BOARD_SLUG" async></script>

Lovable adds the script and your app gets a floating feedback button. Users submit and vote without leaving your product.

To ship what they vote for: copy the top-ranked spec from your board (Generate spec on any request, or this API call) and paste it into Lovable’s chat.

curl -H "Authorization: Bearer fq_YOUR_KEY" https://featq.com/api/v1/top-spec

What it looks like in practice

Shipping a voted-for feature in Lovable looks like this:

You: [paste the FeatQ spec for "Saved filters", your #1 request
  with 19 votes]
  Build this feature exactly as specified.

Lovable: I see the spec: saved filter presets with names, a default
  filter option, and the limit of 10 presets per user from the
  acceptance criteria. Building the UI and storage now...

Lovable: Done. The filter bar now has a save button and a presets
  dropdown.

You: [back on FeatQ, mark "Saved filters" as done. FeatQ publishes
  it to your changelog and emails the 19 voters.]

One quirk to know

Lovable regenerates code as you iterate, and injected snippets can get dropped in big rewrites. If the feedback button disappears after a large change, re-paste the widget prompt; it is idempotent.

Frequently asked questions

Practical answers about agents, voting, embeds, and pricing.

Lovable’s agent builds your app but does not act as a general MCP client you can point at outside servers. The widget plus paste-the-spec flow covers the same loop: collect votes in your app, ship the winners.

Straight to your hosted FeatQ board. The widget opens your board in an overlay, so votes and requests land in the same place your specs come from.

Scope, acceptance criteria, and edge cases pulled from the request and its comment thread, written so a coding agent can build from it directly.

Then you get the native route: connect Claude Code or Cursor to your board over MCP and the agent pulls ranked requests itself. Same board, same key.

Still have questions? Contact us

Keep going

Other clients