← Rize + Vercel integration

Scheduled routine

Vercel MCP: Recap Deployment Work with Rize

Compare available Vercel deployment context with recorded project hours.

A release recap based on actual Rize time, with available deployment references.

How to set it up

1

In Rize, open Settings → Integrations and select the MCP filter. Find Vercel, connect with your own account and approve access. Available tools depend on your account and connector permissions.

2

Open Routines, create a routine, and select Vercel as a connector. Paste the example prompt below and specify the clients, projects and date range you want reviewed.

3

Choose a daily or weekly schedule and your preferred run time. Save and enable the routine. You can also run it manually to check the first result.

4

Review the finished brief in Rize. Check source matches and missing context before making changes. These example prompts request read-only reviews, not unattended writes.

What should your routine ask?

This prompt requests a read-only review. Replace the date range and project scope before you save it.

Use the Vercel connector to read available deployment context for this week. Compare it with list_projects and list_my_time_entries in Rize. Recap related recorded work and flag uncertain project matches. Deployment timestamps are context, not proof of time worked. Do not change deployments or time entries.

Connector tools are discovered when the routine runs. Access depends on your connected account. Routines use AI credits.

Frequently Asked Questions

Open Settings → Integrations, select the MCP filter, and connect Vercel. Then select it in your routine. The routine can use only the connected tools your account permits.

No. This example prompt asks for a review brief without record changes. Routines can run write tools where supported, so review any prompt changes before enabling a schedule.

Check the connection and account permissions, then narrow the prompt to information the connector exposes. Missing context should be flagged in the brief, not filled in with guesses. Scheduled runs use AI credits.

Need help with this integration? Contact support