Developer Tools & Devops · PlanetScale + Claude
PlanetScale MCP with Claude: Safer Database Changes Through Natural Language
The short answer
Claude works with PlanetScale through the Model Context Protocol (MCP). Once connected, Claude can read database branches, deploy requests, backups and production insights, and carry out approved actions in natural language. The organisation controls which permissions Claude has, and people approve consequential changes.
What is PlanetScale?
PlanetScale is a MySQL-compatible database platform built for branching and scalable development. Teams use it to make schema changes safely, using deploy requests to review changes before they reach production, along with backups and production insights to monitor performance. Engineering, data and platform teams are the typical users, often working alongside product and operations staff who depend on the systems these databases support.
How Claude connects to PlanetScale through MCP
The Model Context Protocol (MCP) is an open standard that lets Claude securely use other software, reading information and taking actions with permissions the organisation controls. Think of it as a universal adapter between Claude and the tool. For PlanetScale, the connection is made using OAuth sign-in, so access is granted through the standard PlanetScale sign-in flow rather than shared passwords.
Once connected, Claude can help with tasks such as listing databases, creating branches for development work, reviewing open deploy requests, and inspecting backups and insights. Staff can ask questions in plain English, for example asking which schema changes are waiting for review, and receive a clear summary.
Permissions and approvals keep people in charge. Access is limited to what each role needs, following least-privilege principles. Consequential actions, such as approving a deploy request, can be set to require a named person's sign-off, and every action can be recorded in an audit log for review.
What you can automate
Creating development branches on request
Engineers often need a safe copy of the database to test a new feature. Claude can create a branch from a plain-language request, name it according to your conventions, and report back its status. This reduces waiting time and keeps experiments away from live data.
Summarising pending deploy requests
Engineering leads can ask Claude which schema changes are waiting for review, who raised them, and what they affect. Claude produces a short digest each morning, so reviewers can prioritise their time and nothing sits unreviewed for long.
Investigating backup status
Operations teams can ask Claude whether recent backups completed and when the last restore point was taken. This helps during audits and incident reviews, where quick answers matter and manual checking is slow.
Reviewing production insights after releases
After a release, product and platform teams can ask Claude to highlight unusual query patterns or performance changes in production insights. Claude summarises the findings in plain terms, helping teams decide whether to roll back or investigate further.
Helping new team members learn the workflow
Onboarding staff can ask Claude how branching and deploy requests work in their own environment. Claude can walk them through the steps using their real projects as examples, shortening the time before they contribute confidently.
An end-to-end workflow: Safer schema release with an approval gate
TriggerA developer asks Claude in chat to prepare a schema change for an order-tracking feature.
Claude creates a development branch from the current production schema and reports its name.
The developer applies the schema change on the branch, and Claude summarises the differences for review.
Claude raises a deploy request and posts a summary to the engineering channel in Microsoft 365 Teams.
A named lead reviews and approves the request; Claude records the approval in the ticket for that change in the team's ticketing system.
After deployment, Claude checks production insights and flags any unusual query behaviour to the on-call engineer.
OutcomeSchema changes move from idea to production with a documented review trail, fewer surprises after release, and faster sign-off because reviewers receive clear summaries instead of raw technical output. Results vary by team and process.
How we set it up
Free automation audit
We review how your teams currently manage database changes, backups and releases, and identify the workflows worth automating first. There is no charge for this session.
Secure connection with OAuth sign-in
We connect Claude to PlanetScale using OAuth sign-in, the method PlanetScale supports, so access is granted through your own account and can be revoked at any time.
Workflow design with approvals
We design the end-to-end workflow across PlanetScale and your other systems such as Microsoft 365, your CRM or ticketing tool, and define which actions need a named person's approval.
Testing against real past cases
We test the workflow using real past deploy requests and incidents, checking accuracy and adjusting permissions so Claude only has the access each task requires.
Launch, monitoring and team training
We launch with audit logs and monitoring in place, then train your team on how to use Claude safely and how to review its outputs.
Who benefits
- Engineering manager
- Gets clear, plain-language summaries of pending schema changes and approvals, making review faster and better documented.
- Database or platform engineer
- Spends less time on routine branch creation and status checks, and more time on design and reliability work.
- Operations or IT lead
- Can answer backup and audit questions quickly, with a record of who approved what.
- Product manager
- Can ask how a release affected production in plain terms, without needing to read technical dashboards.
Security and control
- Least-privilege access: Claude only receives the permissions each role needs.
- Named human approval for consequential actions such as approving deploy requests.
- Audit logs of every action Claude takes, available for review and compliance checks.
- Revocable OAuth access, so connections can be ended immediately if needed.
- Testing on non-production branches before any change is applied to live data.
For teams in the US, UK, India and the Middle East
Organisations in the United States and the United Kingdom can host their Claude workloads in those regions, aligning with local expectations for data handling. Indian organisations are subject to the Digital Personal Data Protection Act, 2023, and we design access and logging to support those obligations. Businesses in the UAE and Saudi Arabia must meet local data protection laws, and we choose the hosting region to meet those requirements. Regional compliance depends on your specific obligations and should be confirmed with your legal team.
Learn more about Claude and MCP
Anthropic's AI model and assistant.
Secure links between Claude and business tools.
How engineers build and change software with Claude.
Hosted agents that run multi-step work.
The open standard Claude uses to connect to tools.
Frequently asked questions
What does the PlanetScale MCP server do?
The PlanetScale MCP server acts as a secure bridge between AI tools like Claude and your PlanetScale databases. It lets Claude read information such as branches, deploy requests, backups and insights, and carry out approved actions. It follows the permissions you set, so Claude can only do what you have allowed.
Is connecting Claude to PlanetScale secure?
Security depends on how the connection is configured. We use OAuth sign-in, apply least-privilege access, require named approval for consequential actions and keep audit logs. Access can be revoked at any time. We never guarantee security outcomes, but we design each connection with these controls from the start.
What can Claude actually do with PlanetScale?
Typically, Claude can list databases, create branches, review open deploy requests, and inspect backups and production insights. It can answer questions in plain English and summarise findings. Actions that change live systems are normally set to need a person's approval before they take effect.
How long does setup take?
Timelines vary with the number of workflows and systems involved. A focused setup for a single team often takes a few weeks, including the audit, connection, testing and training. Larger, multi-team rollouts take longer. We confirm the timeline during your free automation audit.
What plan or permissions do I need?
You need a PlanetScale account with the access level that matches the tasks you want Claude to perform, and you sign in using OAuth. Exact plan requirements depend on your PlanetScale subscription, so we confirm what is needed during the audit rather than assume it.