Containerise: Postgres + Redis/RQ + API + Ghidra worker

Brings up the documented target architecture as a docker-compose stack — a
modular monolith with the Ghidra step split into its own async worker.

- worker/: RQ queue (lazy redis import) + run_acquisition task (Job status
  queued→started→finished/failed, drives ams.acquire with sink=db)
- Job model + JobOut schema; Snapshot.data is JSONB on Postgres
- POST/GET /jobs: stream an upload to a shared volume, enqueue, poll status
- docker/api.Dockerfile (slim) + docker/worker.Dockerfile (JDK21 + Ghidra
  fetched at build, overridable via GHIDRA_URL) + docker-compose.yml
- ghidra.py: AMS_GHIDRA_SCRIPTS override for in-container script path
- pyproject: [worker] extra (rq/redis/psycopg), python-multipart in [api]
- tests: 4 new (task success/failure + endpoint enqueue/503) -> 22/22

Verified: API image builds, container serves /health + /ui + /jobs; compose
config validates. Worker image (downloads ~1 GB Ghidra) not built here.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

ams/worker/tasks.py

@@ -0,0 +1,55 @@
+"""The RQ task body. Runs inside the Ghidra-equipped worker container.
+
+It re-points SQLAlchemy at the shared DATABASE_URL (the worker is a separate process
+from the API), drives the acquisition pipeline, and walks the Job row through its
+status transitions so the API/UI can poll progress."""
+
+from __future__ import annotations
+
+import os
+import traceback
+
+from ..acquire import acquire
+from ..api.db import configure, get_session, init_db
+from ..api.models import Job
+
+
+def _set(job_id: int, **fields):
+    """Patch a Job row in its own short-lived session (worker may run far from the API)."""
+    db = get_session()
+    try:
+        job = db.get(Job, job_id)
+        if job is None:
+            return None
+        for k, v in fields.items():
+            setattr(job, k, v)
+        db.commit()
+        return job
+    finally:
+        db.close()
+
+
+def run_acquisition(
+    job_id: int,
+    source_path: str,
+    game_name: str | None = None,
+    database_url: str | None = None,
+) -> dict:
+    """Acquire `source_path` into the catalog, updating Job #`job_id` as it goes.
+
+    Returns a small result dict (also stored by RQ). Errors are recorded on the Job row
+    and re-raised so RQ marks the job failed."""
+    configure(database_url)
+    init_db()
+
+    _set(job_id, status="started", error=None)
+    try:
+        result = acquire(source_path, game_name, sink="db")
+    except Exception as exc:  # record then re-raise so RQ sees the failure too
+        _set(job_id, status="failed", error="{0}: {1}".format(type(exc).__name__, exc))
+        traceback.print_exc()
+        raise
+
+    _set(job_id, status="finished", snapshot_id=result.imported_id,
+         dll_name=os.path.basename(result.dll), error=None)
+    return {"snapshot_id": result.imported_id, "engine": result.engine, "dll": result.dll}