Loading...
You already think out loud with an AI assistant. You draft ideas in Claude, refine them in ChatGPT, or build in Cursor. So why leave that conversation to run market research somewhere else?
The PreVibe MCP connector brings your validation research into the chat you are already in. Ask your assistant to validate an idea, and it runs a real PreVibe report, reads the verdict, and keeps going - all in the same thread. No copy-paste, no tab-switching.
MCP (Model Context Protocol) is an open standard that lets AI apps connect to outside tools. PreVibe ships an MCP server, so any MCP-capable assistant can use PreVibe on your behalf.
Your assistant becomes the front door to PreVibe. You type an idea in plain language, and the report comes back in the same conversation. The context you already built up - your notes, your constraints, your earlier questions - stays right there.
Because the assistant drives the tools, it can string steps together. Validate an idea, then pull a go-to-market deep dive, then ask a follow-up about the riskiest assumption. You describe the goal; it runs the steps.
The assistant is not making the market data up. Each call runs an actual PreVibe report, grounded in real, public sources, and spends a credit like any other report. You get the same verdict and depth you would from the dashboard, delivered inline.
The connector exposes eleven tools. In practice they cover four jobs:
You never call these by name. You just ask, and the assistant picks the right tool.
Connecting used to mean generating an API key and pasting it into a config file. Not anymore.
PreVibe now uses OAuth. You add the server URL to your AI app, sign in with Google, approve access once, and you are connected. There are no keys to copy, store, or rotate. You can see and revoke every connected app from your MCP settings at any time.
The server URL is the same everywhere:
https://previbe.app/api/mcp
Add the server, then authenticate:
claude mcp add --transport http previbe https://previbe.app/api/mcp
Run /mcp inside Claude Code and choose Authenticate. Your browser opens, you sign in with Google, and you are done. Verify with claude mcp list.
Open Settings -> Connectors, click Add custom connector, paste the URL above, then click Connect and sign in. Custom connectors need a Claude Pro, Max, Team, or Enterprise plan. Turn PreVibe tools on from the + menu in any chat.
Add a custom (remote) MCP connector with the same URL. When prompted, sign in with Google and approve access. Your app handles the rest.
The MCP Configuration page always has the current steps and your list of connected apps.
Once connected, try a short chain like this:
Your assistant runs the reports, reads them, and answers - without you leaving the thread.
You sign in with OAuth, so your assistant never sees a password or a long-lived key. A connected app can access your account data - your reports and credit balance - only to act on your behalf, and you can revoke it anytime from your MCP settings. Your use of a connected app is also governed by that app's own terms.
The research itself keeps our usual promise: every AI call runs with zero data retention, so your idea is not stored or trained on by our providers.
If you already have a PreVibe account, connect in about a minute from the MCP Configuration page. New here? Start with Getting Started with Product Research, then bring PreVibe into the assistant you use every day.
Your startup idea is unreleased IP. Learn what zero data retention means, why it matters for founders, and how PreVibe now enforces it on every AI request.
Follow the proven sequence of research methods successful founders use to go from idea to product-market fit.
Maximize the value of your market research with these proven strategies for lean SaaS founders and entrepreneurs.