Connect FeatQ to Gemini CLI
Gemini CLI is Google’s terminal coding agent, and it supports MCP natively. MCP (Model Context Protocol) is an open standard that lets an AI agent call tools on an external server, in this case your FeatQ board. Once connected, Gemini CLI can pull your top-voted requests and build from a generated spec right in your terminal.
Setup: one JSON snippet
Add this to ~/.gemini/settings.json (create the file if it does not exist, or merge the mcpServers block into your existing settings). Replace fq_YOUR_KEY with the key from your board admin page under Agent access.
{
"mcpServers": {
"featq": {
"httpUrl": "https://featq.com/api/mcp",
"headers": {
"Authorization": "Bearer fq_YOUR_KEY"
}
}
}
}Restart Gemini CLI so it picks up the new server. Run /mcp inside the CLI to confirm featq is listed, then ask about your board.
What it looks like in practice
A typical session in Gemini CLI looks like this:
You: What do my users want most on the FeatQ board? Gemini CLI: [calls get_ranked_requests] Top request: "Webhook support" (37 votes, 11 comments). Second: "Custom domains" (24 votes). Ranked by votes plus comments, with recent activity weighted higher. You: Spec the webhooks one and build it. Gemini CLI: [calls generate_spec] Spec covers the endpoint config UI, retry policy, and the payload signing question raised in the comments. Implementing... Gemini CLI: Done, tests pass. Marked the request in progress on the board with update_request_status.
One quirk to know
Gemini CLI distinguishes httpUrl (streamable HTTP) from url (SSE) in its MCP config. The FeatQ server speaks streamable HTTP, so use httpUrl as shown above; with url the connection will not work.
Frequently asked questions
Practical answers about agents, voting, embeds, and pricing.
Yes. Gemini CLI also reads .gemini/settings.json inside a project. Put the server there if you only want FeatQ available in one repo, or in ~/.gemini/settings.json for everywhere.
Six: list_requests, get_request, get_ranked_requests, generate_spec, update_request_status, and get_board_stats. get_ranked_requests sorts by votes plus half-weight comments, with older requests fading over time.
Yes. Gemini CLI runs non-interactively with a prompt argument, so a cron job or GitHub Action can feed it the top spec from the FeatQ API and have it open a PR. Your keys, your schedule, a human reviews the PR.
No. The connection only pulls board data (requests, votes, comments, specs) into the CLI. FeatQ never sees your repository.
Still have questions? Contact us