Developer Tools & Devops · Rootly + Claude
Rootly MCP with Claude: Faster, Calmer Incident Response for Engineering Teams
The short answer
Claude connects to Rootly through the Model Context Protocol, an open standard that lets Claude use software securely with permissions you control. Once connected, Claude can create incidents, update their status, automate response steps and fetch incident summaries from plain-English requests. Access is set up with an API key.
What is Rootly?
Rootly is an AI-powered incident management platform. Engineering, site reliability and operations teams use it to coordinate how they respond when a system fails or slows down. It helps teams manage incident handling step by step, keep everyone informed, and reduce the time that services stay down. It integrates with everyday collaboration tools such as Slack and PagerDuty, so responders can work where they already communicate.
How Claude connects to Rootly through MCP
The Model Context Protocol, or MCP, is an open standard that acts like a universal adapter between Claude and other software. It lets Claude securely read information from a tool and take actions in it, with the permissions the organisation decides on. For Rootly, the connection can run through Rootly's MCP server or its direct API, and access is set up using an API key.
Once connected, Claude can work with Rootly in natural language. For example, a responder can ask Claude to open a new incident with a severity and summary, change an incident's status as work progresses, or pull a clear summary of what has happened so far. Claude can also help trigger the response steps your team has defined, so routine tasks happen consistently.
Your organisation stays in charge. Access is limited to the permissions granted through the API key, and sensitive or consequential actions can be set to require a person's approval before they run. Activity can be logged so that every change to an incident is traceable. The result is faster handling of routine work while people keep control of decisions.
What you can automate
Opening incidents from plain-language reports
When a support lead reports that customers cannot log in, Claude can turn that message into a structured Rootly incident with a summary and severity. The on-call engineer receives a clear starting point without someone re-typing the details. This shortens the gap between a problem being noticed and the right people being engaged.
Keeping incident status current
During a busy outage, status changes are easy to forget. Claude can update an incident's status when a responder reports progress, such as a fix being deployed or monitoring having begun. Stakeholders see accurate information, and the team spends less time on administration and more on resolving the problem.
Fetching incident summaries for leadership
Operations directors and executives often need a quick view of what is happening without reading long threads. Claude can fetch and condense the current state of an incident into a short, readable summary. This helps leaders make decisions faster and keeps non-technical stakeholders informed.
Automating routine response steps
Many incidents follow familiar steps, such as assigning owners, notifying teams and opening follow-up tasks. Claude can help automate these workflow steps through Rootly, so they happen the same way every time. Consistency reduces the chance of a missed step during a stressful incident.
Preparing post-incident reviews
After an incident is resolved, teams need to review what happened and why. Claude can gather the incident's key facts and status history from Rootly and draft a starting outline for the review. Engineers then refine the analysis, which speeds up learning and improves future response.
An end-to-end workflow: Customer-impacting outage to resolution
TriggerA monitoring alert or a support ticket reports that a core service is failing for customers.
Claude reads the alert or ticket and creates a Rootly incident with a draft summary and suggested severity.
Rootly's response steps assign an owner and notify the on-call team through Slack.
As engineers work, they tell Claude about progress, and Claude updates the incident status in Rootly.
When the fix is deployed, Claude records the resolution and the affected services.
A summary is fetched and posted to the relevant ticket in your ticketing system, with a person approving the message before it is sent.
A post-incident review outline is drafted from the incident record for the team to complete.
OutcomeTypical outcomes include a faster start to response, fewer missed updates and a clearer record for the post-incident review. Results vary by organisation and depend on how your processes are set up.
How we set it up
Book a free automation audit
We review how your teams currently handle incidents, from detection through to review, and identify the workflows worth automating with Rootly and Claude.
Create a secure API key
Your administrator generates an API key in Rootly with only the permissions the agreed workflows need. We follow least-privilege access, so Claude can do no more than is required.
Connect Claude to Rootly through MCP
We connect Claude to Rootly using its MCP server or direct API, authenticated with the API key. The connection is tested in a controlled environment before anything touches live incidents.
Design and test the workflow
We build the end-to-end workflow across Rootly and your other systems, such as Slack, PagerDuty and your ticketing tools. We test it against real past incidents and agree which actions require human approval.
Launch, monitor and train your team
We launch with monitoring and audit logs in place, then train responders, on-call engineers and managers on how to work with Claude. We review performance with you after launch and refine the workflow as needed.
Who benefits
- Site reliability and on-call engineers
- Less administrative work during an incident, so more time goes to fixing the problem.
- Engineering and IT managers
- Consistent response steps and a clearer record for reviewing how incidents were handled.
- Operations and support leads
- Quicker, more accurate reporting of incidents so customers and colleagues are kept informed.
- Executives and business leaders
- Short, current summaries that support faster decisions without reading long incident threads.
Security and control
- Access is limited to the permissions granted through the API key, following least-privilege principles.
- Consequential actions can be set to require approval from a named person before they run.
- Activity is logged so every change to an incident can be traced and audited.
- Credentials are stored securely and can be rotated or revoked at any time.
- Workflows are tested against real past cases before launch to reduce the risk of errors.
For teams in the US, UK, India and the Middle East
Organisations in the United States and the United Kingdom have different data protection expectations, and we design access and logging to support them. In India, the Digital Personal Data Protection Act, 2023 shapes how personal data is handled, and workflows are configured accordingly. Organisations in the UAE and Saudi Arabia must meet their own data protection laws, and we help choose a hosting region that meets those requirements for the custom applications we build.
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 Rootly MCP server do?
The Rootly MCP server provides a secure way for AI tools like Claude to work with Rootly using natural language. It can help create incidents, update their status, automate response workflows and fetch incident summaries. Your team decides which actions are permitted, and the server follows the access granted through your API key.
Is connecting Claude to Rootly secure?
Security depends on how the connection is set up. We use an API key with least-privilege permissions, limit what Claude can do, require approval for sensitive actions and keep audit logs. Credentials can be rotated or revoked at any time. We cannot guarantee security outcomes, but we design the setup to reduce risk.
What can Claude do with Rootly once connected?
Typically, Claude can create incidents, update their status, help trigger response steps and fetch summaries of current incidents. It works from plain-English requests, so responders do not need to learn new commands. The exact actions available depend on the permissions granted to the API key.
How long does setup take?
Timelines vary with the number of workflows and systems involved. A focused integration with a few workflows can often be set up within a few weeks, including testing and training. We confirm a realistic timeline during the free automation audit, once we understand your processes.
Which plan or permissions do I need?
You need a Rootly account with an API key that has permissions for the workflows you want to automate. The exact requirements depend on your Rootly setup and the actions involved. Your administrator should confirm this, and we will review it with you during the audit.