Repository navigation
feat(repo): serve the starter site from one worker - #41
Conversation
e7592da to
cc69021
Compare
a6419c4 to
19ec81c
Compare
5bf36be to
fe8620d
Compare
fe8620d to
7240056
Compare
3f59428 to
c12182c
Compare
7240056 to
77c8175
Compare
77c8175 to
c078b77
Compare
c078b77 to
2b763c1
Compare
|
Fixes |
4ce06fb to
fb53998
Compare
2b763c1 to
ab9a728
Compare
ab9a728 to
7e9744a
Compare
fb53998 to
4f8ea78
Compare
7e9744a to
892ff74
Compare
892ff74 to
aa48fd5
Compare
aa48fd5 to
f346bd8
Compare
68974f4 to
bbe39c1
Compare
f68c69c to
0f0e072
Compare
starter-verify F10 and F13 (Kiro ruling, cycle 35): drop workflow_dispatch so mutation runs on push to main only; the README no longer claims Cell workflows, Cell.provide, tenant-bound store ports or compile-time phase markers, and no longer tells readers to run pnpm mutation locally
Ryan's standing rule via Kiro (cycle 52): no macOS anywhere. The flake builds x86_64-linux and aarch64-linux only; dprint pins only Linux archives; the comment-checker sandbox loses its sandbox-exec branch; AGENTS.md says CI is Linux only
# Conflicts: # AGENTS.md
Conductor ruling, cycle 62: the same bump pnpm-release-management#31 took. effect and every @effect/* stay at 4.0.1; none resolves lower. tsgo 0.50.0 reports nothing new: typecheck and lint are clean uncached. @systemfsoftware/oxlint-config-recommended 4.0.0 still pulls its own @effect/tsgo 0.45.0
# Conflicts: # pnpm-lock.yaml
# Conflicts: # pnpm-lock.yaml # pnpm-workspace.yaml
There was a problem hiding this comment.
Conductor verdict: all checks green on ca40e7f, 0 threads, no unchecked boxes, hunt grep clean.
Lake 1, U6, stacked on #37. Trimmed per Kiro's 2026-10-06 ruling on the starter spec.
apps/site(@endgame/site) is the one Cloudflare Worker: a TanStack Start site defined with Alchemy 2 (Cloudflare.Website.Viteinalchemy.run.ts, entrysrc/worker.ts, compatibility date and flags insite-worker.ts).pnpm --filter @endgame/site devruns it underalchemy dev, Alchemy's local emulation in workerd, with local state only.apps/site-e2eholds the end-to-end journeys; its one script,journeys, runs them againstSITE_URL(defaulthttp://localhost:1337).apps/site: the home route (src/routes/index.tsx), the root route, the router, andvite-plugins.ts. The official@cloudflare/vite-pluginbuilds the Worker forpnpm build; inside Alchemy's build child (ALCHEMY_CLOUDFLARE_VITE_INJECTED=1) it is skipped, because Alchemy injects its own and a second one broke every deploy.apps/site-e2e:home.integration.test.ts(the home page is server-rendered HTML carrying the starter heading) andrequest-timeout.integration.test.ts(a request that never answers is abandoned at its 30 s deadline withSiteRequestTimedOut).@cloudflare/vite-plugin, alchemy 2.0.0-beta.80, andworkerdpinned through an override so one version resolves.Removed by the trim: Markdown/HTML negotiation,
/llms.txt, the README-opening home page and its Vite plugin, the agent journey, the front-door taxonomy andtrace-taxonomy, the workerd render-failure harness, and the local/deployed vitest split.Gate at
3aa8168