how Sorus fits into a Rust workflow
The relationship between your workflow, cloud execution and a Sorus job.
For: Rust developers evaluating Sorus
Sorus runs Rust checks, builds and tests in the cloud. A terminal, CI system or coding agent submits the work and receives a job identifier. The job connects the captured source and requested command to its logs, result and selected artifacts.
execution and orchestration
A CI system decides when to run a workflow and which steps depend on others. Sorus executes the supported Rust job within that workflow. Checkout, packaging, deployment and non-Rust tasks can remain in the existing CI system.
This separation lets a team evaluate one Rust job without moving the entire pipeline. Terminal and agent callers use the same execution service without requiring GitHub Actions.
get started
The Cargo check tutorial teaches a first submission. If you already know how Sorus jobs work, go directly to CI integration, agent setup or the CLI reference.
Rust CI basics explains Cargo operations and continuous integration for readers new to CI. The team evaluation guide covers choosing a representative company workload.
beta access
Onboarding supplies the client and confirms the release, authorised repositories, machine classes and terms. Apply for access before attempting a customer submission.
These docs describe the source revision reviewed on 6 October 2026. The support matrix lists candidate workloads still undergoing release qualification. Availability depends on the beta release and account.
job lifecycle
A job moves through source capture and upload, admission, execution and result delivery. Authentication identifies the caller; repository authorisation and workload checks decide whether the request can be admitted.
Once admitted, the job has an identity independent of the client connection. Disconnecting can leave it running. The client can follow that same job again or retrieve its selected artifacts before expiry.
A compiler or test failure reports something about the requested workload. A platform failure reports a failure to execute or observe that work. CI integrations need this distinction when deciding whether to retry, fall back or report a failed check. The job recovery guide shows the corresponding actions.