wehub-resource-sync
33 lines
1.5 KiB
TypeScript
33 lines
1.5 KiB
TypeScript
import type { ProbeTarget } from "./verify-deploy";
|
|
import type { ProbeOutcome } from "./verify-deploy.drivers";
|
|
import { probeBaseline } from "./verify-deploy.drivers.baseline";
|
|
|
|
/**
|
|
* Feature-level verifier for the starter-template container fleet
|
|
* (`starter-<slug>`). Each starter ships a SINGLE deployable image that
|
|
* EXPOSEs the Next.js frontend on port 3000; that frontend serves `/` and
|
|
* `/api/copilotkit` but has NO `/api/health` route (the agent's `/health`
|
|
* lives on the internal agent port 8123, which Railway does not expose). So
|
|
* `/` is the only correct, reachable healthcheck for the exposed surface —
|
|
* the same baseline the Next.js shells use, and the same `healthcheckPath`
|
|
* the provisioner (`provision-starter-fleet.ts`) sets on the Railway
|
|
* instance config.
|
|
*
|
|
* Like every driver this composes on `probeBaseline`, which enforces the
|
|
* two minimum invariants before any feature-level extension:
|
|
* 1. Railway deployment-SUCCESS for the resolved env (GraphQL).
|
|
* 2. HTTP 200 on `https://<host>/`.
|
|
*
|
|
* Driver-specific DOM + network-call extensions (e.g. a CopilotKit
|
|
* round-trip against `/api/copilotkit`) can compose on top of this baseline
|
|
* once those micro-tasks land; until then this baseline is the agreed
|
|
* completion bar, and crucially NO driver remains a fail-loud "not handled"
|
|
* stub.
|
|
*/
|
|
export async function probeStarter(target: ProbeTarget): Promise<ProbeOutcome> {
|
|
return probeBaseline(target, {
|
|
driverLabel: "starter",
|
|
healthcheckPath: "/",
|
|
});
|
|
}
|