Run a scoped remote task and hand off tested branches or pull requests.
Coding & development · Documented · Documentation checked 2026-10-04
Cursor editor agents and Cloud Agents; Grok Bot has a separate shared-computer boundary.
Installed client, subscription, organization settings, repository access, and cloud environment determine capabilities.
An illustrative recipe. This website does not run the automation.
The goal and the right kind of system
Run a scoped remote task and hand off tested branches or pull requests.
Cloud Agents fit prepared remote development. A fixed review / CI process governs acceptance; agent completion does not deploy the result.
Inputs and connected services
Inputs: approved source connection, prepared VM environment, task criteria, and reviewer.
Services: Cursor Cloud Agents and the authorized source provider; optional test-scoped MCP tools only.
The steps
An illustrative sequence to adapt and test before using real data or enabling ongoing work.
- Inspect repo privileges and environment setup.
- Start the scoped cloud task and confirm source / branch.
- Inspect dependencies, tool activity, and actual checks.
- Review the diff and handoff before any merge.
Outputs and human checkpoints
Outputs: reviewable branch / PR and evidence of checks and blockers.
Human checkpoints: approve environment authority and review the produced change.
Exceptions, permissions, and stopping conditions
Permission boundary: approved repos and test services; no production account actions.
Stop on unavailable required services, failed verification, unexpected multi-repo scope, or secrets / network requirements not reviewed.
Adaptable prompt
Complete [scoped task] in the prepared cloud environment. Verify branch, dependencies, and actual checks. Produce a reviewable handoff. Stop on missing services or broader authority; no merge, deployment, or production changes.
Related reading
Cloud Agents and development environments
Prepare a remote repository workspace and review changes before merging.
Plugins, MCP, and permission boundaries
Evaluate external tools and command approvals before extending an agent.
Review, recovery, and operational limits
Use diffs, checks, and checkpoints as complementary evidence rather than automatic acceptance.
Designing an AI workflow
Make a useful task repeatable with clear inputs, steps, outputs, exceptions, and human checkpoints.
Permissions and meaningful approvals
Decide what a system may read, what it may change, and when someone needs to authorize a concrete action.
Sources and verification
Documentation checked 2026-10-04. Examples are illustrative unless described otherwise.
Cloud Agents and environments · Cursor
Agent security and approvals · Cursor
Agent fundamentals and checkpoints · Cursor