Nuxt AI Chatbot

nuxt-ai-chatbot

Vercel's Nuxt AI Chatbot — the Nuxt UI chat template: an AI chat with streaming answers, reasoning, chart and weather tools, chat history, file uploads and GitHub sign-in — running on Cloudflare Workers + Static Assets through Nitro's cloudflare_module preset, with its models on Workers AI through the Worker's own binding, its chats in a Cloudflare D1 database and its uploads in an R2 bucket, both provisioned by Terraform through Orun and the schema applied by migrations, and packaged as an Orun baseline: one command takes a new chat from nothing to live on stage and prod, in its own repository, under its own name — in ONE landing. Workers Free is enough: the Worker is about 1.2 MiB gzipped.

The template's code is the code. NuxtHub keeps its hub:db and hub:blob with Cloudflare drivers — D1 in place of Turso, R2 in place of Vercel Blob — and the port changes what the Workers runtime and an unattended deployment force: the models default to Workers AI, so a fresh product chats with no model key (the template's own AI Gateway models appear once its key is set, as in ai-chatbot); the schema is migrations applied before the Worker deploys; the session key is generated in the database; uploads are served back by the Worker; and GitHub sign-in waits, visibly, for its owner to connect it. What the baseline adds is everything needed to make a copy of it somebody else's, built the way turso-starter is. See specs/epics/nuxt-ai-chatbot for the design, the milestones and the as-built status.

Instantiating a product

git clone --depth 1 --branch <baseline-tag> https://github.com/sourceplane/nuxt-ai-chatbot
cd nuxt-ai-chatbot
orun new --blueprint repo-blueprint.yaml --out ~/sites/acme-chat --run-hooks --resume \
  --set reponame=acme-chat --set productname="Acme Chat" \
  --set githuborg=acme-inc --set subdomain=<workers-dev-subdomain>

Or, once the platform's registry lists it, from the Orun console (Baselines → nuxt-ai-chatbot → Build). Two phases: 01-product creates the repository and lands the whole product as ONE pull request — intent, CI, tooling, the D1 and R2 roots, the migrations, the chat and its tests — and one convergence, per environment, applies the database and the bucket, then the migrations, then deploys the chat with their names; 02-docs records the live deployment from probed facts. BOOTSTRAP.md is the operator guide; docs/phases/ has a page per phase.

The new repository receives the product only — the app, its D1 and R2 roots, its migrations and their runner, its tests, its CI, its tooling, a README of its own and its operating contract — renamed throughout to the product's name, repository, Cloudflare Worker, database and bucket names. None of this repository's machinery ships. No AI vendor key is asked for: the platform brokers none, and the chat runs on Workers AI.

What is here (the product)

Getting started

pnpm install
pnpm run wire:fixture   # render apps/web/wrangler.jsonc from the offline fixture
pnpm typecheck
pnpm lint
pnpm test
pnpm build              # nuxt build → apps/web/.output (the Worker and its static assets)

pnpm --filter @app/web db:migrate:local   # the schema and the generated session key in wrangler's local D1
pnpm --filter @app/web preview            # wrangler dev over the built Worker, on the local D1 and R2
pnpm --filter @app/web dev                # nuxt dev — the bindings through wrangler's platform proxy

Workers AI has no local mode: both reach it remotely, so wrangler login first.

How it maps to Vercel

How the database and the bucket reach the Worker

infra/terraform/cloudflare-d1 ──apply──► orun secret WIRING_CLOUDFLARE_D1 (stage, prod)
infra/terraform/cloudflare-r2 ──apply──► orun secret WIRING_CLOUDFLARE_R2 (stage, prod)
                                               │
                                               ▼
infra/db-migrate  dependsOn cloudflare-d1 (enforced on main) · secretEnv WIRING_CLOUDFLARE_D1
  apply:  packages/db's runner → POST …/d1/database/<id>/query, each pending migration once
                                               ▼
apps/web  dependsOn cloudflare-d1 + cloudflare-r2 + db-migrate (enforced on main)
          · secretEnv WIRING_CLOUDFLARE_D1_<ENV>, WIRING_CLOUDFLARE_R2_<ENV>
          · optionalSecretEnv the owner's GitHub OAuth app, AI Gateway key, session password
  build:  tooling/wire/live.mjs → wrangler.jsonc with the environment's database id and bucket
          (dev, verify-only, renders the fixture) → nuxt build (without the optional secrets)
          → tooling/wire/worker-secrets.mjs → `wrangler secret bulk` (deploy lanes only)

wrangler.jsonc is committed as the fixture render of wrangler.template.jsonc, because the deploy lane's verify-app-structure step checks the config exists before it builds; testing/e2e.test.sh fails if the two disagree.

CI