Files
accounted/Dockerfile
T
Mattsson 11995b1b0c feat(auth): make automatic logout an opt-in per-user setting (#1536)
* feat(auth): make automatic logout an opt-in per-user setting

Session timeouts (30 min idle / 12 h absolute on hosted) now apply only
to users who enable "Automatic logout" in Settings > Security. Default
is off: sessions live for the full Supabase refresh-token lifetime, the
behavior from before the 2026-07 session hardening.

- user_preferences.auto_logout (migration, default false), toggled via
  the extended /api/user/preferences route
- The opt-in is snapshotted into the signed timeout cookie at mint, so
  enforcement stays DB-read-free per request; the preferences route
  clears the cookie on change so a toggle takes effect immediately
- Pre-toggle cookies are authentic-but-stale: re-minted preserving
  their timers, never routed down the tamper path, so the rollout does
  not log anyone out
- NEXT_PUBLIC_SESSION_TIMEOUT_FORCE_ALL=true enforces timeouts for
  every user regardless of preference (emergency lever, also plumbed
  through the Docker image); self-hosted stays disabled by default

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(auth): resolve PR #1536 review findings

- Replace the spread upsert in /api/user/preferences with one literal
  payload per field: the phantom-column schema guard cannot resolve
  spread payloads (Unit tests 3/4 ceiling failure)
- Map the preferences 500 through getErrorMessage so the user-facing
  text is Swedish (CodeRabbit)
- fetchAutoLogoutPreference now returns null on a FAILED read instead
  of a fail-open false: callers skip minting so an unknown preference
  is never persisted into the year-long signed cookie, and the next
  request retries; failures log at error level, distinct from the
  normal opt-out path (compliance swarm GDPR Art.32(1)(b) / ISO A.8.5)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(auth): write multi-field preference updates as one atomic upsert

A request carrying both hide_assistant_fab and auto_logout previously
issued two sequential writes, so a failure of the second returned 500
after half the request had persisted (CodeRabbit, PR #1536). One
literal upsert per accepted field combination keeps the write atomic
and stays resolvable for the phantom-column schema guard.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-08-12 16:45:41 +02:00

108 lines
5.1 KiB
Docker

# ── Stage 1: Base ──
FROM node:22-alpine@sha256:968df39aedcea65eeb078fb336ed7191baf48f972b4479711397108be0966920 AS base
# `apk upgrade` patches OS packages (e.g. libssl3/libcrypto3) that have fixes
# published after the pinned base digest was built, so the Trivy image scan in
# CI doesn't fail on fixable Alpine CVEs. The digest stays pinned for a
# reproducible starting point; only security patches float on top.
RUN apk upgrade --no-cache && apk add --no-cache libc6-compat
# ── Stage 2: Dependencies ──
FROM base AS deps
WORKDIR /app
COPY package.json package-lock.json ./
RUN npm ci
# ── Stage 3: Builder ──
FROM base AS builder
WORKDIR /app
ARG EXTENSIONS_PRESET=self-hosted
COPY --from=deps /app/node_modules ./node_modules
COPY . .
# Apply extension preset (must happen before build: prebuild hook
# runs setup:extensions which reads extensions.config.json)
COPY docker/extensions.${EXTENSIONS_PRESET}.json ./extensions.config.json
# Build with placeholder sentinel values for NEXT_PUBLIC_* vars.
# These get replaced at runtime by docker-entrypoint.sh so the image
# is generic and reusable across different Supabase projects.
ENV NEXT_PUBLIC_SUPABASE_URL=__NEXT_PUBLIC_SUPABASE_URL__
# Realtime WebSocket origin for the CSP. Must be its own sentinel: the CSP is
# baked into the build output, so the entrypoint cannot derive wss:// from the
# already-substituted https URL after the fact (issue #893).
ENV NEXT_PUBLIC_SUPABASE_WS_URL=__NEXT_PUBLIC_SUPABASE_WS_URL__
ENV NEXT_PUBLIC_SUPABASE_ANON_KEY=__NEXT_PUBLIC_SUPABASE_ANON_KEY__
ENV NEXT_PUBLIC_APP_URL=__NEXT_PUBLIC_APP_URL__
ENV NEXT_PUBLIC_VAPID_PUBLIC_KEY=__NEXT_PUBLIC_VAPID_PUBLIC_KEY__
ENV NEXT_PUBLIC_SELF_HOSTED=__NEXT_PUBLIC_SELF_HOSTED__
ENV NEXT_PUBLIC_REQUIRE_MFA=__NEXT_PUBLIC_REQUIRE_MFA__
ENV NEXT_PUBLIC_SESSION_IDLE_TIMEOUT_MS=__NEXT_PUBLIC_SESSION_IDLE_TIMEOUT_MS__
ENV NEXT_PUBLIC_SESSION_ABSOLUTE_TIMEOUT_MS=__NEXT_PUBLIC_SESSION_ABSOLUTE_TIMEOUT_MS__
ENV NEXT_PUBLIC_SESSION_WARNING_MS=__NEXT_PUBLIC_SESSION_WARNING_MS__
ENV NEXT_PUBLIC_SESSION_TIMEOUT_FORCE_ALL=__NEXT_PUBLIC_SESSION_TIMEOUT_FORCE_ALL__
# Keep the branding placeholder intact through prebuild's inject script so
# docker-entrypoint.sh can substitute the runtime value into public/sw.js.
ENV NEXT_PUBLIC_BRANDING_APP_NAME=__NEXT_PUBLIC_BRANDING_APP_NAME__
ENV NEXT_TELEMETRY_DISABLED=1
RUN npm run build
# ── Stage 4: Runner ──
FROM node:22-alpine@sha256:968df39aedcea65eeb078fb336ed7191baf48f972b4479711397108be0966920 AS runner
WORKDIR /app
# Patch OS packages (libssl3/libcrypto3, …) with fixes published after the
# pinned base digest, so CI's Trivy scan doesn't flag fixable Alpine CVEs. No
# su-exec or curl needed: the entrypoint runs unprivileged as nextjs and the
# healthcheck uses BusyBox wget. The runtime runs `node server.js` and never
# invokes npm, so we delete the base image's bundled npm CLI: its vendored deps
# (picomatch, tar, brace-expansion, ip-address) are the packages Trivy flags on
# this image: removing npm clears them at the source and shrinks the attack
# surface.
RUN apk upgrade --no-cache && \
rm -rf /usr/local/lib/node_modules/npm /usr/local/bin/npm /usr/local/bin/npx
ENV NODE_ENV=production
ENV NEXT_TELEMETRY_DISABLED=1
RUN addgroup --system --gid 1001 nodejs && \
adduser --system --uid 1001 -G nodejs nextjs
# /app at runtime is split across the read-only image layer and tmpfs mounts:
# /app/server.js, /app/node_modules/, /app/package.json : image (read-only)
# /app/.next/ : tmpfs (writable)
# /app/public/ : tmpfs (writable)
# The entrypoint runs UNPRIVILEGED as nextjs: it copies the templates from
# /opt/gnubok-template/ into the nextjs-owned tmpfs mounts, substitutes the
# NEXT_PUBLIC_* placeholders, then drops the write bits. Because it never needs
# to chown or setuid, the container runs under docker-compose's `cap_drop: ALL`
# + `read_only: true` with no added capabilities.
COPY --from=builder --chown=nextjs:nodejs /app/.next/standalone/server.js ./server.js
COPY --from=builder --chown=nextjs:nodejs /app/.next/standalone/node_modules ./node_modules
COPY --from=builder --chown=nextjs:nodejs /app/.next/standalone/package.json ./package.json
# Baked-in templates for runtime population of the tmpfs mounts.
COPY --from=builder --chown=nextjs:nodejs /app/.next/standalone/.next /opt/gnubok-template/.next
COPY --from=builder --chown=nextjs:nodejs /app/.next/static /opt/gnubok-template/.next/static
COPY --from=builder --chown=nextjs:nodejs /app/public /opt/gnubok-template/public
# Pre-create the tmpfs mount points (empty in the image layer; the entrypoint
# fills them at startup). Owned by nextjs so the unprivileged entrypoint can
# write into the tmpfs mounted over them.
RUN mkdir -p /app/.next/cache /app/public && \
chown nextjs:nodejs /app /app/.next /app/.next/cache /app/public
COPY --chmod=755 --chown=nextjs:nodejs docker-entrypoint.sh ./docker-entrypoint.sh
USER nextjs
EXPOSE 3000
ENV PORT=3000
ENV HOSTNAME=0.0.0.0
ENTRYPOINT ["./docker-entrypoint.sh"]
CMD ["node", "server.js"]