mass deploy billing-api --env devOne environment. Multiple interfaces.
A developer can deploy from the CLI. An agent can query dependencies and propose a change. Automation can use the API. Platform teams can inspect changes, approvals, and dependencies through the UI.They are not separate delivery systems. They are different interfaces to the same environment.
CLIConcept
MCP serverToday
what depends on orders-db?GraphQL APIToday
deployments, instances, approvalsUIToday
graph · approvals · change historyThe same commands for you and your agent
Output is structured and every command maps to the API, so a coding agent can drive the CLI or the MCP server without screen scraping. It gets the same plan, the same policy checks, and the same approval gates you do.
A coding agent asks what depends on orders-db, then proposes an upgrade that goes through plan, policy, and approval
agent> what depends on orders-db in prod? orders-db aws-rds-postgres@1.3 ├── billing-api ├── worker └── reporting-job agent> propose upgrading orders-db to 1.4 plan 2 changes, 0 destroys impact 3 dependents, connection params unchanged policy ✓ passed (prod-db-change) approval required from @platform → proposal #418 waiting for review
From an empty account to a running app
Create a project, add a database, link it to your app, deploy, and promote.
Creating a billing project with dev and prod environments
$ mass project create billing
✓ project billing created
$ mass environment create billing-{dev,prod}
✓ billing-dev created
✓ billing-prod created