Security review for the pull requests Cursor's cloud agent opens.
Cursor's cloud agent builds and tests the change in its own virtual machine, then hands you a pull request. Aevral, the AI security reviewer for GitHub pull requests, reviews that pull request within your plan's allowance and gives you a copy-paste fix prompt with each finding.
In plain words
Picture assigning a task to a contractor who works somewhere else entirely: Cursor's cloud agent is built for that. It runs in its own isolated virtual machine in the cloud, with a full development environment, builds and tests the change there, and can open the result as a pull request on GitHub. This page is about that pull request.
How Cursor's cloud agent gets to a pull request
From Cursor's own documentation. Each line is in our words, with the quote it rests on underneath.
Cloud agents run in isolated VMs in the cloud with a full development environment, instead of on your own machine.
“Cloud agents use the same agent fundamentals but run in isolated VMs in the cloud with full development environments instead of on your local machine.”
The docs describe cloud agent as a background agent that typically runs for a long time.
“Background agents are coding agents that typically run for a very long time.”
It clones your GitHub repository to work on the change.
“Cloud agents clone your repo from GitHub”
Because it has its own virtual machine, it can build, test, and interact with the changed software before handing it back.
“Because they have access to their own virtual machine, cloud agents can build, test, and interact with the changed software.”
The same agent can build features, fix bugs, write tests, and open the pull request automatically when it finishes.
“These agents can build features, fix bugs, write tests, and automatically open pull requests when they finish.”
Cursor Help, Background Agents (opening pull requests), read
A comment that starts with @cursor on a GitHub pull request or issue can start a cloud agent task.
“Comment @cursor on a GitHub PR or issue”
Starting one needs read-write privileges to the repository and any dependent repositories or submodules.
“You need read-write privileges to your repo and any dependent repos or submodules.”
Cloud agent runs remotely, apart from your local editor
You can run several cloud agents in parallel, without keeping your own machine connected to the internet.
“in parallel, and they do not require your local machine to be connected to the internet.”
Because it runs remotely, you can close your laptop and check the result later.
“Because it runs remotely, you can close your laptop and check the result later.”
Sending the agent's URL to a teammate lets them open the run to see the conversation, the code changes, and the artifacts it produced, without pulling the branch locally.
“Send an agent's URL to a teammate and they can open the run to see the conversation, the code changes, and the artifacts the agent produced.”
Why it matters: the account that opens it, and what the artifacts don't cover
Cursor's cloud agent does not always open the pull request under the same name.
The docs say an automation that runs as you opens the pull request as your own GitHub account, and one that runs as the shared service account opens it as cursor instead.
Either way, the diff is what a reviewer, human or Aevral, has to read; which account opened the pull request does not decide what shipped.
The docs describe a merge-ready pull request, with screenshots, videos and logs attached so you can validate the change. That evidence is about whether the feature works, not whether it is safe.
Aevral reads the diff on that same pull request when it is opened, within your plan's review allowance, looking for access control, business logic and injection problems among its classes, and leaves its finding on the added line where the reviewer is already looking.
What the vendor's docs say
An automation that runs as you opens the pull request as your own GitHub account.
“Automations that run as you open pull requests as your GitHub account.”
One that runs as the shared service account opens it as cursor instead.
“Automations that run as the service account open pull requests as cursor”
GitHub comments, review approvals and reviewer requests from the service account show up as cursor too.
“GitHub comments, review approvals, and reviewer requests run as cursor”
The docs call the result a merge-ready pull request, with artifacts attached to demo the change.
“Cloud agents produce merge-ready PRs with artifacts to demo their changes.”
As it works, it also attaches videos, screenshots and logs to the pull request so the change can be checked without pulling the branch.
“As an agent works, it tests its changes and attaches videos, screenshots, and logs to the pull request, so you can validate the work without checking out the branch.”
Where Aevral fits: a copy-paste handoff
Aevral reviews a Cursor cloud agent pull request like any other pull request on a repository where the Aevral GitHub App is installed: when it is opened or reopened and when new commits are pushed, within your plan's review allowance.
A finding is an inline comment on the added line, next to an advisory Check that never blocks the pull request.
Each finding carries a Fix with your agent prompt.
To hand it to the cloud agent, a human copies the prompt and pastes it into a comment that starts with @cursor on the pull request, or into a new cloud agent task.
What comes back is reviewed again when it reaches a pull request, and your repository's own review and merge rules decide what ships.
How the prompt is built and how to use it well: handing a finding to your coding agent.
Setup, the two human clicks
Two clicks, both yours: install the GitHub App on the repositories the agent works in, then log in to the console with GitHub. An agent cannot install the App or finish the login for you; it can read this page and the agent-setup prompt instead.
- Install the GitHub App: https://github.com/apps/aevral
- Log in to the console: https://app.aevral.com/login
- Official agent-setup prompt: https://docs.aevral.com/agent-setup/prompt.md
What this page does not claim
- Aevral is not affiliated with, endorsed by, or a partner of the company behind Cursor. There is no integration between the two products: Aevral does not call Cursor, and Cursor does not call Aevral.
- Aevral never commits, pushes, applies or merges a change. The Fix with your agent prompt is text a human copies; what the agent does with it goes through your repository's own review and merge rules.
- PR review looks for a list of classes on the diff and leaves at most five findings per review. A finding is a lead with evidence, not a verdict, and a quiet review is not proof that the code is safe.
- The boxed statements about Cursor come from the vendor's own documentation, each with the quote it rests on and the date it was read. The paragraphs around them are Aevral's reading of what those facts mean for a security review. Vendor products change; the vendor's docs win over this page.
- This page does not claim that Aevral is a Cursor Bugbot competitor and does not compare or rank the products.
- Aevral does not start cloud agent tasks, read the agent's virtual machine, or act under the cursor service account.
Sources
Read next