Managed agents
Delegate the work. Choose where it runs.
Run delegated engineering tasks in configured agent sandboxes, alongside local execution. Agent execution is separate from where the platform and underlying model are hosted.
LatentStack is a modular AI software stack. This component is available on its own.
How it works
01
Configured execution sandboxes
02
Goals and reviewable outputs
03
Execution and inference are separate
Illustrative managed-agent task
Works with
Execution
Configured sandboxes and local execution, selected per task.
Models
Whatever LatentGateway permits for that workload; no separate model policy applies here.
Repositories
Git-based code hosts, with task output returned as reviewable changes.
Supported-capability list v0.1 · owned by engineering · requires engineering sign-off before publication.
What it does not do
−It does not run unattended delivery. Outputs are returned for review.
−It does not change where inference happens. Execution location and model hosting are set separately.
−Consumption terms for managed execution are not yet defined; they are agreed during scoping.
Deployment note
Sandboxes are configured per deployment. In an air-gapped configuration, execution and inference both stay inside the isolated environment.
Works alone. Better with
Questions
What does it cost?
Platform pricing is per seat; underlying model usage is paid by the customer to the provider. Component-level packaging and consumption terms for managed execution are set during scoping.
Where does it run?
Managed by LatentForce, on your infrastructure, or in an isolated air-gapped environment.