build(docker): slim backend image, static frontend image, -docling variant

Backend (arc53/docsgpt): 4.5 GB compressed -> 0.9 GB with both embedding
models and tiktoken baked in.
- torch/transformers gone from the default install (docling extra only).
- Ubuntu 24.04 ships python3.12: no deadsnakes PPA, no software-properties-
  common; every pin is a wheel, so no gcc/g++/rust in the builder.
- COPY --chown and a prefetch that runs as the process user replace the
  trailing chown -R, which duplicated the 600 MB model layer.
- .dockerignore keeps __pycache__, .coverage, local indexes and .env out.
- EXTRAS build arg (INSTALL_DOCLING kept as an alias); the docling variant
  also bakes docling's layout/table/RapidOCR models (DOCLING_ARTIFACTS_PATH)
  and tesseract, and drops only the discovery documents of Google APIs the
  app never builds.
- FLASK_DEBUG env removed (unused); OCI labels added.

Frontend (arc53/docsgpt-fe): 302 MB Vite dev server -> 25 MB static build
behind nginx. VITE_* variables are injected at container start into
/config.js and read through src/env.ts, so the image no longer needs a
rebuild per deployment; docker-compose.yaml keeps hot reload via the dev
target.

Publishing: every release and develop build now pushes a slim tag and a
-docling tag (docling engine + models + tesseract). docker-compose-hub.yaml
takes DOCSGPT_IMAGE_TAG / DOCSGPT_IMAGE_VARIANT; docker-compose-standalone.yaml
runs the stack from pre-built images without a checkout and is attached to
each release. setup.sh selects the -docling variant for OCR instead of
requiring a local build. A new workflow builds the image on PRs that touch
it and runs verify_offline under --network none; lint checks the exported
requirements match uv.lock.

deployment/docker-compose-azure.yaml

@@ -1,6 +1,8 @@
 services:
   frontend:
-    build: ../frontend
+    build:
+      context: ../frontend
+      target: dev
     environment:
       - VITE_API_HOST=http://localhost:7091
       - VITE_API_STREAMING=$VITE_API_STREAMING
@@ -13,6 +15,7 @@ services:
     build:
       context: ../application
       args:
+        EXTRAS: ${EXTRAS:-}
         INSTALL_DOCLING: ${INSTALL_DOCLING:-false}
         # Off by default; deployments running OCR_ENABLED=true with tesseract
         # must set INSTALL_TESSERACT=true before rebuilding (see docker-compose.yaml).
@@ -40,6 +43,7 @@ services:
     build:
       context: ../application
       args:
+        EXTRAS: ${EXTRAS:-}
         INSTALL_DOCLING: ${INSTALL_DOCLING:-false}
         # Off by default; deployments running OCR_ENABLED=true with tesseract
         # must set INSTALL_TESSERACT=true before rebuilding (see docker-compose.yaml).

deployment/docker-compose-hub.yaml

@@ -1,8 +1,14 @@
+# Pre-built images from Docker Hub (mirrored at ghcr.io/arc53).
+#   DOCSGPT_IMAGE_TAG      develop (default, follows main) or a release, e.g. 0.20.0
+#   DOCSGPT_IMAGE_VARIANT  empty (default, slim) or -docling: docling parser engine,
+#                          its models, and tesseract baked in (OCR-ready)
+# Set them in ../.env or the shell. deployment/docker-compose-standalone.yaml is
+# the same stack without a git checkout.
 name: docsgpt-oss
 services:
 
   frontend:
-    image: arc53/docsgpt-fe:develop
+    image: arc53/docsgpt-fe:${DOCSGPT_IMAGE_TAG:-develop}
     environment:
       - VITE_API_HOST=http://localhost:7091
       - VITE_API_STREAMING=${VITE_API_STREAMING:-true}
@@ -15,7 +21,7 @@ services:
 
   backend:
     user: root
-    image: arc53/docsgpt:develop
+    image: arc53/docsgpt:${DOCSGPT_IMAGE_TAG:-develop}${DOCSGPT_IMAGE_VARIANT:-}
     env_file:
       - ../.env
     environment:
@@ -38,7 +44,7 @@ services:
 
   worker:
     user: root
-    image: arc53/docsgpt:develop
+    image: arc53/docsgpt:${DOCSGPT_IMAGE_TAG:-develop}${DOCSGPT_IMAGE_VARIANT:-}
     # `parsing` queue carries read_document/parse_document; required for its await to resolve.
     command: celery -A application.app.celery worker -l INFO -B -Q docsgpt,parsing,embeddings
     env_file:

deployment/docker-compose-standalone.yaml

@@ -0,0 +1,104 @@
+# DocsGPT from pre-built images, with no git checkout.
+#
+#   curl -fsSLO https://raw.githubusercontent.com/arc53/DocsGPT/main/deployment/docker-compose-standalone.yaml
+#   printf 'LLM_PROVIDER=docsgpt\nVITE_API_STREAMING=true\n' > .env   # or any provider, see the settings guide
+#   docker compose up -d
+#   open http://localhost:5173
+#
+# Every release also attaches this file as an asset. Settings come from .env
+# next to this file (any DocsGPT setting; the compose-internal service URLs
+# below take precedence). Data lives in named volumes, so `docker compose
+# down` keeps it and `docker compose down -v` removes it.
+#
+#   DOCSGPT_IMAGE_TAG      release to run, e.g. 0.20.0 (default: latest release);
+#                          develop follows the main branch
+#   DOCSGPT_IMAGE_VARIANT  empty (slim, default) or -docling: docling parser
+#                          engine, its models, and tesseract baked in (OCR-ready)
+name: docsgpt
+
+services:
+  frontend:
+    image: arc53/docsgpt-fe:${DOCSGPT_IMAGE_TAG:-latest}
+    env_file:
+      - path: .env
+        required: false
+    environment:
+      - VITE_API_HOST=${VITE_API_HOST:-http://localhost:7091}
+      - VITE_API_STREAMING=${VITE_API_STREAMING:-true}
+    ports:
+      - "5173:5173"
+    depends_on:
+      - backend
+
+  backend:
+    image: arc53/docsgpt:${DOCSGPT_IMAGE_TAG:-latest}${DOCSGPT_IMAGE_VARIANT:-}
+    env_file:
+      - path: .env
+        required: false
+    environment:
+      - CELERY_BROKER_URL=redis://redis:6379/0
+      - CELERY_RESULT_BACKEND=redis://redis:6379/1
+      - CACHE_REDIS_URL=redis://redis:6379/2
+      - POSTGRES_URI=postgresql://docsgpt:docsgpt@postgres:5432/docsgpt
+    ports:
+      - "7091:7091"
+    volumes:
+      - indexes:/app/indexes
+      - inputs:/app/inputs
+      - vectors:/app/vectors
+    depends_on:
+      redis:
+        condition: service_started
+      postgres:
+        condition: service_healthy
+    restart: unless-stopped
+
+  worker:
+    image: arc53/docsgpt:${DOCSGPT_IMAGE_TAG:-latest}${DOCSGPT_IMAGE_VARIANT:-}
+    # Consumes the default queue plus `parsing` (read_document) and `embeddings`
+    # (query embedding); without the latter every search times out.
+    command: celery -A application.app.celery worker -l INFO -B -Q docsgpt,parsing,embeddings
+    env_file:
+      - path: .env
+        required: false
+    environment:
+      - CELERY_BROKER_URL=redis://redis:6379/0
+      - CELERY_RESULT_BACKEND=redis://redis:6379/1
+      - CACHE_REDIS_URL=redis://redis:6379/2
+      - POSTGRES_URI=postgresql://docsgpt:docsgpt@postgres:5432/docsgpt
+      - API_URL=http://backend:7091
+    volumes:
+      - indexes:/app/indexes
+      - inputs:/app/inputs
+      - vectors:/app/vectors
+    depends_on:
+      redis:
+        condition: service_started
+      postgres:
+        condition: service_healthy
+    restart: unless-stopped
+
+  redis:
+    image: redis:6-alpine
+    restart: unless-stopped
+
+  postgres:
+    image: postgres:16-alpine
+    environment:
+      - POSTGRES_USER=docsgpt
+      - POSTGRES_PASSWORD=docsgpt
+      - POSTGRES_DB=docsgpt
+    volumes:
+      - postgres_data:/var/lib/postgresql/data
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U docsgpt -d docsgpt"]
+      interval: 5s
+      timeout: 5s
+      retries: 10
+    restart: unless-stopped
+
+volumes:
+  indexes:
+  inputs:
+  vectors:
+  postgres_data:

deployment/docker-compose.yaml

@@ -1,7 +1,11 @@
 name: docsgpt-oss
 services:
   frontend:
-    build: ../frontend
+    build:
+      context: ../frontend
+      # Vite dev server with hot reload over the bind mount below. The default
+      # target (what Docker Hub publishes) is a static build behind nginx.
+      target: dev
     volumes:
       - ../frontend/src:/app/src
     environment:
@@ -18,8 +22,10 @@ services:
     build:
       context: ../application
       args:
-        # Bake the optional docling engine (layout-model OCR backend, structured
-        # output) into the image: set INSTALL_DOCLING=true in ../.env or the shell.
+        # Optional extras to bake in (comma-separated): docling, milvus. The
+        # docling extra brings the layout-model parser/OCR backend and its
+        # models. INSTALL_DOCLING=true is the older spelling of EXTRAS=docling.
+        EXTRAS: ${EXTRAS:-}
         INSTALL_DOCLING: ${INSTALL_DOCLING:-false}
         # Bake the tesseract binary behind OCR_ENABLED=true (~35 MB): set
         # INSTALL_TESSERACT=true in ../.env or the shell (setup.sh does this
@@ -53,6 +59,7 @@ services:
     build:
       context: ../application
       args:
+        EXTRAS: ${EXTRAS:-}
         INSTALL_DOCLING: ${INSTALL_DOCLING:-false}
         INSTALL_TESSERACT: ${INSTALL_TESSERACT:-false}
     # Consumes the default queue AND the dedicated `parsing` (read_document /