# Lyra Walking Skeleton Implementation Plan
> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
**Goal:** Stand up the end-to-end Lyra architecture — a Next.js UI/API, a Python worker, and Postgres — where a user submits an album request, the worker claims the resulting job and drives it through the full state machine via a *fake* pipeline, and the UI shows live progress. Zero source integration.
**Architecture:** Three Docker Compose services — `web` (Next.js App Router, owns the Prisma schema), `worker` (Python poll-loop service), and `db` (Postgres). The web app writes `Request` + `Job` rows; the worker claims jobs with `SELECT … FOR UPDATE SKIP LOCKED` and advances each through the state machine. Communication is entirely through Postgres, exactly as slices 2–4 will use it.
**Tech Stack:** Next.js 15 (App Router, TypeScript, React 19), Prisma 6, PostgreSQL 17, Python 3.12 with psycopg 3. Tests: Vitest 2 (web API), pytest 8 (worker).
## Global Constraints
- Node.js 22 LTS; Next.js App Router (not Pages Router); TypeScript strict mode on.
- Python 3.12 in the worker container (`python:3.12-slim`); local test runs may use a newer 3.x — keep worker deps version-flexible so they install on both. Database access via `psycopg` v3 (not psycopg2); raw SQL only in the worker.
- **Prisma is the single owner of the database schema.** The worker never issues DDL; it reads/writes tables Prisma migrated. Worker tests assume `prisma migrate deploy` has already been run against the target database.
- PostgreSQL 17.
- One `DATABASE_URL` (Postgres connection string) is shared by both services via environment.
- Job state values (exact): `requested`, `matching`, `matched`, `downloading`, `tagging`, `imported`, `needs_attention`.
- Request status values (exact): `pending`, `completed`, `needs_attention`.
- Monorepo layout: `web/` (Next.js), `worker/` (Python), root holds `docker-compose.yml` and `.env.example`.
- Every task ends with a commit.
---
### Task 1: Monorepo scaffold + Docker Compose
**Files:**
- Create: `web/package.json`
- Create: `web/tsconfig.json`
- Create: `web/next.config.ts`
- Create: `web/src/app/layout.tsx`
- Create: `web/src/app/page.tsx`
- Create: `worker/requirements.txt`
- Create: `worker/lyra_worker/__init__.py`
- Create: `worker/tests/__init__.py`
- Create: `worker/pytest.ini`
- Create: `docker-compose.yml`
- Create: `.env.example`
**Interfaces:**
- Consumes: nothing (first task).
- Produces: a runnable Next.js app at `web/`, an importable Python package `lyra_worker`, and a Compose stack with services `web`, `worker`, `db` sharing `DATABASE_URL`.
- [ ] **Step 1: Create the Next.js app manifest and config**
`web/package.json`:
```json
{
"name": "lyra-web",
"version": "0.1.0",
"private": true,
"scripts": {
"dev": "next dev",
"build": "next build",
"start": "next start",
"test": "vitest run"
},
"dependencies": {
"next": "15.1.0",
"react": "19.0.0",
"react-dom": "19.0.0",
"@prisma/client": "6.1.0"
},
"devDependencies": {
"typescript": "5.7.2",
"@types/node": "22.10.2",
"@types/react": "19.0.1",
"@types/react-dom": "19.0.1",
"prisma": "6.1.0",
"vitest": "2.1.8"
}
}
```
`web/tsconfig.json`:
```json
{
"compilerOptions": {
"target": "ES2022",
"lib": ["dom", "dom.iterable", "ES2022"],
"allowJs": true,
"skipLibCheck": true,
"strict": true,
"noEmit": true,
"esModuleInterop": true,
"module": "esnext",
"moduleResolution": "bundler",
"resolveJsonModule": true,
"isolatedModules": true,
"jsx": "preserve",
"incremental": true,
"plugins": [{ "name": "next" }],
"paths": { "@/*": ["./src/*"] }
},
"include": ["next-env.d.ts", "**/*.ts", "**/*.tsx", ".next/types/**/*.ts"],
"exclude": ["node_modules"]
}
```
`web/next.config.ts`:
```ts
import type { NextConfig } from "next";
const nextConfig: NextConfig = {
output: "standalone",
};
export default nextConfig;
```
- [ ] **Step 2: Create a minimal placeholder UI so the app builds**
`web/src/app/layout.tsx`:
```tsx
export const metadata = { title: "Lyra" };
export default function RootLayout({ children }: { children: React.ReactNode }) {
return (
{children}
);
}
```
`web/src/app/page.tsx`:
```tsx
export default function Home() {
return Lyra;
}
```
- [ ] **Step 3: Create the Python worker package skeleton**
`worker/requirements.txt`:
```
psycopg[binary]>=3.2,<4
pytest>=8.3,<9
```
`worker/lyra_worker/__init__.py`:
```python
"""Lyra acquisition worker."""
__version__ = "0.1.0"
```
`worker/tests/__init__.py`:
```python
```
`worker/pytest.ini`:
```ini
[pytest]
testpaths = tests
```
- [ ] **Step 4: Create the Compose stack and env template**
`.env.example`:
```
POSTGRES_USER=lyra
POSTGRES_PASSWORD=lyra
POSTGRES_DB=lyra
DATABASE_URL=postgresql://lyra:lyra@db:5432/lyra
```
`docker-compose.yml`:
```yaml
services:
db:
image: postgres:17
environment:
POSTGRES_USER: ${POSTGRES_USER}
POSTGRES_PASSWORD: ${POSTGRES_PASSWORD}
POSTGRES_DB: ${POSTGRES_DB}
volumes:
- postgres-data:/var/lib/postgresql/data
healthcheck:
test: ["CMD-SHELL", "pg_isready -U ${POSTGRES_USER} -d ${POSTGRES_DB}"]
interval: 5s
timeout: 5s
retries: 5
web:
build: ./web
environment:
DATABASE_URL: ${DATABASE_URL}
ports:
- "3000:3000"
depends_on:
db:
condition: service_healthy
worker:
build: ./worker
environment:
DATABASE_URL: ${DATABASE_URL}
depends_on:
db:
condition: service_healthy
volumes:
postgres-data:
```
- [ ] **Step 5: Verify the scaffold builds**
Run:
```bash
cp .env.example .env
cd web && npm install && npm run build
cd ../worker && python -m pip install -r requirements.txt && python -c "import lyra_worker; print(lyra_worker.__version__)"
```
Expected: `next build` completes without errors; the Python import prints `0.1.0`.
- [ ] **Step 6: Verify Compose config is valid**
Run: `docker compose config -q && echo OK`
Expected: prints `OK` (no schema errors).
- [ ] **Step 7: Commit**
```bash
git add web worker docker-compose.yml .env.example
git commit -m "chore: scaffold web, worker, and compose stack"
```
---
### Task 2: Prisma schema + migration for Request and Job
**Files:**
- Create: `web/prisma/schema.prisma`
- Create: `web/src/lib/db.ts`
- Modify: `web/package.json` (add `postinstall` + `db:migrate` scripts)
**Interfaces:**
- Consumes: `DATABASE_URL` from Task 1.
- Produces:
- Tables `Request(id, artist, album, status, createdAt)` and `Job(id, requestId, state, currentStage, attempts, error, claimedAt, createdAt, updatedAt)`.
- Enums `RequestStatus { pending, completed, needs_attention }` and `JobState { requested, matching, matched, downloading, tagging, imported, needs_attention }`.
- A Prisma client singleton exported as `prisma` from `@/lib/db`.
- [ ] **Step 1: Write the Prisma schema**
`web/prisma/schema.prisma`:
```prisma
generator client {
provider = "prisma-client-js"
}
datasource db {
provider = "postgresql"
url = env("DATABASE_URL")
}
enum RequestStatus {
pending
completed
needs_attention
}
enum JobState {
requested
matching
matched
downloading
tagging
imported
needs_attention
}
model Request {
id String @id @default(cuid())
artist String
album String
status RequestStatus @default(pending)
createdAt DateTime @default(now())
job Job?
}
model Job {
id String @id @default(cuid())
request Request @relation(fields: [requestId], references: [id], onDelete: Cascade)
requestId String @unique
state JobState @default(requested)
currentStage String @default("intake")
attempts Int @default(0)
error String?
claimedAt DateTime?
createdAt DateTime @default(now())
updatedAt DateTime @updatedAt
}
```
- [ ] **Step 2: Add the Prisma client singleton**
`web/src/lib/db.ts`:
```ts
import { PrismaClient } from "@prisma/client";
const globalForPrisma = globalThis as unknown as { prisma?: PrismaClient };
export const prisma = globalForPrisma.prisma ?? new PrismaClient();
if (process.env.NODE_ENV !== "production") globalForPrisma.prisma = prisma;
```
- [ ] **Step 3: Add migration/generate scripts to package.json**
In `web/package.json`, add to `"scripts"`:
```json
"postinstall": "prisma generate",
"db:migrate": "prisma migrate deploy",
"db:dev": "prisma migrate dev"
```
- [ ] **Step 4: Create the migration and apply it**
Run (Postgres must be up — `docker compose up -d db`):
```bash
cd web
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
npx prisma migrate dev --name init_request_job
```
Expected: a migration is created under `web/prisma/migrations/` and applied; Prisma prints "Your database is now in sync with your schema."
- [ ] **Step 5: Verify the tables exist**
Run:
```bash
cd web
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
npx prisma db execute --stdin <<'SQL'
SELECT 1 FROM "Request" LIMIT 0;
SELECT 1 FROM "Job" LIMIT 0;
SQL
echo "tables OK"
```
Expected: no error; prints `tables OK`.
- [ ] **Step 6: Commit**
```bash
git add web/prisma web/src/lib/db.ts web/package.json
git commit -m "feat: add Request and Job schema with migration"
```
---
### Task 3: Request API (create + list)
**Files:**
- Create: `web/src/app/api/requests/route.ts`
- Create: `web/vitest.config.ts`
- Create: `web/src/test/setup.ts`
- Test: `web/src/app/api/requests/route.test.ts`
- Modify: `web/package.json` (add `@vitejs/plugin-react` not needed; Vitest config uses node env)
**Interfaces:**
- Consumes: `prisma` from `@/lib/db` (Task 2).
- Produces:
- `POST /api/requests` — body `{ artist: string, album: string }` → creates a `Request` (status `pending`) and its `Job` (state `requested`) in one transaction; responds `201` with `{ id, artist, album, status, job: { id, state, currentStage } }`.
- `GET /api/requests` — responds `200` with `{ requests: Array<{ id, artist, album, status, createdAt, job: { state, currentStage } | null }> }`, newest first.
- [ ] **Step 1: Write the failing test**
`web/vitest.config.ts`:
```ts
import { defineConfig } from "vitest/config";
import { fileURLToPath } from "node:url";
export default defineConfig({
resolve: {
alias: {
"@": fileURLToPath(new URL("./src", import.meta.url)),
},
},
test: {
environment: "node",
setupFiles: ["./src/test/setup.ts"],
fileParallelism: false,
},
});
```
`web/src/test/setup.ts`:
```ts
import { beforeEach } from "vitest";
import { prisma } from "@/lib/db";
beforeEach(async () => {
// Job has a cascade FK to Request; delete Job first.
await prisma.job.deleteMany();
await prisma.request.deleteMany();
});
```
`web/src/app/api/requests/route.test.ts`:
```ts
import { describe, it, expect } from "vitest";
import { POST, GET } from "./route";
function postReq(body: unknown) {
return new Request("http://localhost/api/requests", {
method: "POST",
headers: { "content-type": "application/json" },
body: JSON.stringify(body),
});
}
describe("requests API", () => {
it("creates a request with an initial job", async () => {
const res = await POST(postReq({ artist: "Radiohead", album: "In Rainbows" }));
expect(res.status).toBe(201);
const body = await res.json();
expect(body.artist).toBe("Radiohead");
expect(body.status).toBe("pending");
expect(body.job.state).toBe("requested");
expect(body.job.currentStage).toBe("intake");
});
it("rejects a request missing fields", async () => {
const res = await POST(postReq({ artist: "Radiohead" }));
expect(res.status).toBe(400);
});
it("lists requests newest first with job state", async () => {
await POST(postReq({ artist: "A", album: "1" }));
await POST(postReq({ artist: "B", album: "2" }));
const res = await GET();
expect(res.status).toBe(200);
const body = await res.json();
expect(body.requests).toHaveLength(2);
expect(body.requests[0].artist).toBe("B");
expect(body.requests[0].job.state).toBe("requested");
});
});
```
- [ ] **Step 2: Run the test to verify it fails**
Run:
```bash
cd web
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
npx vitest run src/app/api/requests/route.test.ts
```
Expected: FAIL — cannot import `./route` (module does not exist).
- [ ] **Step 3: Implement the route handlers**
`web/src/app/api/requests/route.ts`:
```ts
import { prisma } from "@/lib/db";
export async function POST(request: Request) {
let body: unknown;
try {
body = await request.json();
} catch {
return Response.json({ error: "invalid JSON" }, { status: 400 });
}
const { artist, album } = (body ?? {}) as { artist?: unknown; album?: unknown };
if (typeof artist !== "string" || !artist.trim() || typeof album !== "string" || !album.trim()) {
return Response.json({ error: "artist and album are required" }, { status: 400 });
}
const created = await prisma.request.create({
data: {
artist: artist.trim(),
album: album.trim(),
job: { create: {} },
},
include: { job: true },
});
return Response.json(
{
id: created.id,
artist: created.artist,
album: created.album,
status: created.status,
job: { id: created.job!.id, state: created.job!.state, currentStage: created.job!.currentStage },
},
{ status: 201 },
);
}
export async function GET() {
const requests = await prisma.request.findMany({
orderBy: { createdAt: "desc" },
include: { job: true },
});
return Response.json({
requests: requests.map((r) => ({
id: r.id,
artist: r.artist,
album: r.album,
status: r.status,
createdAt: r.createdAt,
job: r.job ? { state: r.job.state, currentStage: r.job.currentStage } : null,
})),
});
}
```
- [ ] **Step 4: Run the test to verify it passes**
Run:
```bash
cd web
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
npx vitest run src/app/api/requests/route.test.ts
```
Expected: PASS — all three tests green.
- [ ] **Step 5: Commit**
```bash
git add web/src/app/api web/vitest.config.ts web/src/test/setup.ts
git commit -m "feat: add request create/list API"
```
---
### Task 4: Request form + live queue UI
**Files:**
- Modify: `web/src/app/page.tsx`
- Create: `web/src/app/queue.tsx`
**Interfaces:**
- Consumes: `POST /api/requests` and `GET /api/requests` (Task 3).
- Produces: a client page that submits a new request and polls the queue every 2 seconds, rendering each request's artist/album and live job state.
- [ ] **Step 1: Write the client queue component**
`web/src/app/queue.tsx`:
```tsx
"use client";
import { useEffect, useState } from "react";
type Row = {
id: string;
artist: string;
album: string;
status: string;
job: { state: string; currentStage: string } | null;
};
export function Queue() {
const [rows, setRows] = useState([]);
const [artist, setArtist] = useState("");
const [album, setAlbum] = useState("");
async function refresh() {
const res = await fetch("/api/requests");
const body = await res.json();
setRows(body.requests);
}
useEffect(() => {
refresh();
const id = setInterval(refresh, 2000);
return () => clearInterval(id);
}, []);
async function submit(e: React.FormEvent) {
e.preventDefault();
if (!artist.trim() || !album.trim()) return;
await fetch("/api/requests", {
method: "POST",
headers: { "content-type": "application/json" },
body: JSON.stringify({ artist, album }),
});
setArtist("");
setAlbum("");
refresh();
}
return (
{rows.map((r) => (
-
{r.artist} — {r.album} · {r.job?.state ?? r.status}
{r.job ? ` (${r.job.currentStage})` : ""}
))}
);
}
```
- [ ] **Step 2: Render the queue on the home page**
`web/src/app/page.tsx`:
```tsx
import { Queue } from "./queue";
export default function Home() {
return (
Lyra
);
}
```
- [ ] **Step 3: Verify the app builds**
Run: `cd web && npm run build`
Expected: build completes; `/` is compiled as a route with no type errors.
- [ ] **Step 4: Commit**
```bash
git add web/src/app/page.tsx web/src/app/queue.tsx
git commit -m "feat: add request form and live queue UI"
```
---
### Task 5: Worker job-claim logic
**Files:**
- Create: `worker/lyra_worker/db.py`
- Create: `worker/lyra_worker/claim.py`
- Create: `worker/tests/conftest.py`
- Test: `worker/tests/test_claim.py`
**Interfaces:**
- Consumes: `DATABASE_URL`; the `Job`/`Request` tables from Task 2.
- Produces:
- `connect() -> psycopg.Connection` in `db.py`.
- `claim_next(conn) -> str | None` in `claim.py`: atomically selects the oldest `Job` in state `requested` using `FOR UPDATE SKIP LOCKED`, sets its state to `matching`, `claimedAt = now()`, `currentStage = 'match'`, increments `attempts`, commits, and returns the job id (or `None` if none available).
- [ ] **Step 1: Add the DB connection helper**
`worker/lyra_worker/db.py`:
```python
import os
import psycopg
def connect() -> psycopg.Connection:
dsn = os.environ["DATABASE_URL"]
return psycopg.connect(dsn)
```
- [ ] **Step 2: Write the test fixtures**
`worker/tests/conftest.py`:
```python
import os
import psycopg
import pytest
@pytest.fixture()
def conn():
dsn = os.environ["DATABASE_URL"]
connection = psycopg.connect(dsn)
yield connection
# Clean up rows created during the test (Job cascades from Request).
with connection.cursor() as cur:
cur.execute('DELETE FROM "Job"')
cur.execute('DELETE FROM "Request"')
connection.commit()
connection.close()
def insert_request(conn, artist="Artist", album="Album"):
"""Insert a Request + its Job (state 'requested') and return the job id."""
with conn.cursor() as cur:
cur.execute(
'INSERT INTO "Request" (id, artist, album, status, "createdAt") '
"VALUES (gen_random_uuid()::text, %s, %s, 'pending', now()) RETURNING id",
(artist, album),
)
request_id = cur.fetchone()[0]
cur.execute(
'INSERT INTO "Job" (id, "requestId", state, "currentStage", attempts, "createdAt", "updatedAt") '
"VALUES (gen_random_uuid()::text, %s, 'requested', 'intake', 0, now(), now()) RETURNING id",
(request_id,),
)
job_id = cur.fetchone()[0]
conn.commit()
return job_id
```
- [ ] **Step 3: Write the failing test**
`worker/tests/test_claim.py`:
```python
from lyra_worker.claim import claim_next
from tests.conftest import insert_request
def test_claim_returns_none_when_empty(conn):
assert claim_next(conn) is None
def test_claim_advances_job_to_matching(conn):
job_id = insert_request(conn)
claimed = claim_next(conn)
assert claimed == job_id
with conn.cursor() as cur:
cur.execute('SELECT state, "currentStage", attempts, "claimedAt" FROM "Job" WHERE id = %s', (job_id,))
state, stage, attempts, claimed_at = cur.fetchone()
assert state == "matching"
assert stage == "match"
assert attempts == 1
assert claimed_at is not None
def test_claim_does_not_repick_claimed_job(conn):
insert_request(conn)
assert claim_next(conn) is not None
assert claim_next(conn) is None
```
- [ ] **Step 4: Run the test to verify it fails**
Run:
```bash
cd worker
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
python -m pytest tests/test_claim.py -v
```
Expected: FAIL — `ModuleNotFoundError: No module named 'lyra_worker.claim'`.
- [ ] **Step 5: Implement claim_next**
`worker/lyra_worker/claim.py`:
```python
import psycopg
def claim_next(conn: psycopg.Connection) -> str | None:
"""Atomically claim the oldest queued job. Returns its id or None."""
with conn.cursor() as cur:
cur.execute(
"""
SELECT id FROM "Job"
WHERE state = 'requested'
ORDER BY "createdAt" ASC
FOR UPDATE SKIP LOCKED
LIMIT 1
"""
)
row = cur.fetchone()
if row is None:
conn.rollback()
return None
job_id = row[0]
cur.execute(
"""
UPDATE "Job"
SET state = 'matching',
"currentStage" = 'match',
"claimedAt" = now(),
attempts = attempts + 1,
"updatedAt" = now()
WHERE id = %s
""",
(job_id,),
)
conn.commit()
return job_id
```
- [ ] **Step 6: Run the test to verify it passes**
Run:
```bash
cd worker
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
python -m pytest tests/test_claim.py -v
```
Expected: PASS — all three tests green.
- [ ] **Step 7: Commit**
```bash
git add worker/lyra_worker/db.py worker/lyra_worker/claim.py worker/tests/conftest.py worker/tests/test_claim.py
git commit -m "feat: add worker job-claim logic"
```
---
### Task 6: Fake pipeline through the state machine
**Files:**
- Create: `worker/lyra_worker/pipeline.py`
- Test: `worker/tests/test_pipeline.py`
**Interfaces:**
- Consumes: a claimed job id (state `matching`) from Task 5; the DB connection.
- Produces: `run_pipeline(conn, job_id, stage_delay=0.0) -> None`: advances the job through `matched → downloading → tagging → imported`, updating `state` and `currentStage` at each step, then sets the parent `Request.status` to `completed`. `stage_delay` seconds are slept between stages (default `0.0` for tests; the runtime loop passes a small value so progress is visible in the UI).
- [ ] **Step 1: Write the failing test**
`worker/tests/test_pipeline.py`:
```python
from lyra_worker.claim import claim_next
from lyra_worker.pipeline import run_pipeline
from tests.conftest import insert_request
def test_pipeline_drives_job_to_imported(conn):
job_id = insert_request(conn)
claim_next(conn)
run_pipeline(conn, job_id, stage_delay=0.0)
with conn.cursor() as cur:
cur.execute('SELECT state, "currentStage" FROM "Job" WHERE id = %s', (job_id,))
state, stage = cur.fetchone()
cur.execute('SELECT status FROM "Request" WHERE id = (SELECT "requestId" FROM "Job" WHERE id = %s)', (job_id,))
req_status = cur.fetchone()[0]
assert state == "imported"
assert stage == "import"
assert req_status == "completed"
```
- [ ] **Step 2: Run the test to verify it fails**
Run:
```bash
cd worker
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
python -m pytest tests/test_pipeline.py -v
```
Expected: FAIL — `ModuleNotFoundError: No module named 'lyra_worker.pipeline'`.
- [ ] **Step 3: Implement the fake pipeline**
`worker/lyra_worker/pipeline.py`:
```python
import time
import psycopg
# (state, currentStage) pairs the job moves through, in order, after claim.
_STAGES = [
("matched", "rank"),
("downloading", "download"),
("tagging", "tag"),
("imported", "import"),
]
def _set_state(conn: psycopg.Connection, job_id: str, state: str, stage: str) -> None:
with conn.cursor() as cur:
cur.execute(
'UPDATE "Job" SET state = %s, "currentStage" = %s, "updatedAt" = now() WHERE id = %s',
(state, stage, job_id),
)
conn.commit()
def run_pipeline(conn: psycopg.Connection, job_id: str, stage_delay: float = 0.0) -> None:
"""Fake acquisition: walk the claimed job through to 'imported'."""
for state, stage in _STAGES:
if stage_delay:
time.sleep(stage_delay)
_set_state(conn, job_id, state, stage)
with conn.cursor() as cur:
cur.execute(
'UPDATE "Request" SET status = \'completed\' '
'WHERE id = (SELECT "requestId" FROM "Job" WHERE id = %s)',
(job_id,),
)
conn.commit()
```
- [ ] **Step 4: Run the test to verify it passes**
Run:
```bash
cd worker
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
python -m pytest tests/test_pipeline.py -v
```
Expected: PASS.
- [ ] **Step 5: Run the whole worker suite**
Run:
```bash
cd worker
export DATABASE_URL=postgresql://lyra:lyra@localhost:5432/lyra
python -m pytest -v
```
Expected: PASS — claim and pipeline tests all green.
- [ ] **Step 6: Commit**
```bash
git add worker/lyra_worker/pipeline.py worker/tests/test_pipeline.py
git commit -m "feat: add fake acquisition pipeline"
```
---
### Task 7: Worker main loop + Dockerfiles + end-to-end verification
**Files:**
- Create: `worker/lyra_worker/main.py`
- Create: `worker/Dockerfile`
- Create: `web/Dockerfile`
- Create: `web/.dockerignore`
- Create: `worker/.dockerignore`
- Create: `web/docker-entrypoint.sh`
**Interfaces:**
- Consumes: `claim_next` (Task 5), `run_pipeline` (Task 6), `connect` (Task 5).
- Produces: a long-running worker process (`python -m lyra_worker.main`) that repeatedly claims a job and runs the pipeline, sleeping when idle; production Dockerfiles for both services; the web container runs `prisma migrate deploy` before starting.
- [ ] **Step 1: Implement the worker main loop**
`worker/lyra_worker/main.py`:
```python
import time
from lyra_worker.claim import claim_next
from lyra_worker.db import connect
from lyra_worker.pipeline import run_pipeline
IDLE_SLEEP = 2.0
STAGE_DELAY = 1.0 # visible progress in the UI
def run_forever() -> None:
conn = connect()
print("worker: waiting for jobs", flush=True)
try:
while True:
job_id = claim_next(conn)
if job_id is None:
time.sleep(IDLE_SLEEP)
continue
print(f"worker: claimed job {job_id}", flush=True)
run_pipeline(conn, job_id, stage_delay=STAGE_DELAY)
print(f"worker: imported job {job_id}", flush=True)
finally:
conn.close()
if __name__ == "__main__":
run_forever()
```
- [ ] **Step 2: Add the worker Dockerfile and dockerignore**
`worker/Dockerfile`:
```dockerfile
FROM python:3.12-slim
WORKDIR /app
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt
COPY lyra_worker ./lyra_worker
CMD ["python", "-m", "lyra_worker.main"]
```
`worker/.dockerignore`:
```
tests
__pycache__
*.pyc
```
- [ ] **Step 3: Create the empty public dir and web dockerignore**
Create `web/public/.gitkeep` (empty file — the standalone image copies `public/`, which must exist):
```
```
`web/.dockerignore`:
```
node_modules
.next
.env
```
- [ ] **Step 4: Add the web Dockerfile and entrypoint**
`web/docker-entrypoint.sh`:
```bash
#!/bin/sh
set -e
npx prisma migrate deploy
exec node server.js
```
`web/Dockerfile`:
```dockerfile
FROM node:22-slim AS build
WORKDIR /app
RUN apt-get update && apt-get install -y openssl && rm -rf /var/lib/apt/lists/*
COPY package.json package-lock.json* ./
RUN npm install
COPY . .
RUN npx prisma generate && npm run build
FROM node:22-slim AS run
WORKDIR /app
RUN apt-get update && apt-get install -y openssl && rm -rf /var/lib/apt/lists/*
ENV NODE_ENV=production
COPY --from=build /app/.next/standalone ./
COPY --from=build /app/.next/static ./.next/static
COPY --from=build /app/public ./public
COPY --from=build /app/prisma ./prisma
COPY --from=build /app/node_modules/.bin/prisma ./node_modules/.bin/prisma
COPY --from=build /app/node_modules/prisma ./node_modules/prisma
COPY --from=build /app/node_modules/@prisma ./node_modules/@prisma
COPY --from=build /app/docker-entrypoint.sh ./
RUN chmod +x docker-entrypoint.sh
EXPOSE 3000
CMD ["./docker-entrypoint.sh"]
```
- [ ] **Step 5: Build and start the whole stack**
Run:
```bash
docker compose build
docker compose up -d
sleep 15
docker compose logs worker | tail -n 5
```
Expected: worker logs show `worker: waiting for jobs`; web and db containers are healthy.
- [ ] **Step 6: Verify end-to-end via the API**
Run:
```bash
curl -s -X POST http://localhost:3000/api/requests \
-H "content-type: application/json" \
-d '{"artist":"Radiohead","album":"In Rainbows"}' >/dev/null
# Poll until the request completes (worker takes ~4s with STAGE_DELAY=1s).
for i in $(seq 1 15); do
sleep 2
state=$(curl -s http://localhost:3000/api/requests | python3 -c 'import sys,json;print(json.load(sys.stdin)["requests"][0]["job"]["state"])')
echo "state=$state"
[ "$state" = "imported" ] && break
done
```
Expected: the printed state advances through `matching`/`downloading`/etc. and reaches `imported`.
- [ ] **Step 7: Tear down**
Run: `docker compose down`
Expected: all containers stopped and removed.
- [ ] **Step 8: Commit**
```bash
git add worker/lyra_worker/main.py worker/Dockerfile worker/.dockerignore \
web/Dockerfile web/.dockerignore web/docker-entrypoint.sh web/public/.gitkeep
git commit -m "feat: add worker loop and production Dockerfiles"
```
---
## Notes for the next plan (Pipeline core)
The fake `run_pipeline` in `worker/lyra_worker/pipeline.py` is the seam the next plan replaces. Plan 2 introduces the real six-stage pipeline, the `SourceAdapter` contract, the ranker + confidence scoring, and fake adapters — swapping the fake stage-walk for real stage functions while keeping the claim loop, state machine, and DB schema unchanged. The `candidates`, `releases`, `artists`, `library_items`, and `config` tables from the design spec are added by their respective plans as the stages that need them arrive.