feat(studio): bundle studio-expose bridge + socat (opt-in STUDIO_EXPOSE)

pi-studio binds the container's 127.0.0.1, which a published Docker port can't reach. Add a robust, portable bridge rather than a doc-only one-liner: - Dockerfile.base: add socat (~1 MB, generally useful TCP relay). - rootfs/usr/local/bin/studio-expose: socat TCP relay listening on the container's egress IPv4 (not 0.0.0.0 — that would EADDRINUSE against Studio's loopback listener) forwarding to 127.0.0.1:PORT on the SAME port, so Studio's printed token URL works verbatim. Robust egress-IP detection (hostname -I, loopback-filtered; ip route get fallback), --help, port validation, foreground. - entrypoint-user.sh: opt-in STUDIO_EXPOSE=1 auto-starts the bridge in the background (studio variant only). Default OFF — Studio stays loopback-only (its secure default) unless explicitly opted in. - README: 'Using pi-studio' now documents host-networking (A) and the studio-expose/STUDIO_EXPOSE bridge (B) with a security note; ssh -L for remote, mosh caveat retained. - smoke-test: assert socat + studio-expose present (base-level). - CHANGELOG/AGENTS updated. No tag — stopping for review.
2026-06-10 23:33:44 +02:00
parent a78e59fb5b
commit 7d8ee4cea1
7 changed files with 131 additions and 12 deletions
@@ -120,7 +120,8 @@ deprecated artifacts (to be removed in opencode-devbox v2.0.0).
  `pi install /opt/pi-studio` (see Dockerfile.variant `INSTALL_STUDIO`).
  The default `:latest` image stays studio-free. Note: pi-studio binds
  `127.0.0.1` inside the container, so browser access needs host
-  networking or a loopback bridge — see README "Using pi-studio".
+  networking or the bundled `studio-expose` bridge (socat; auto-starts
+  when `STUDIO_EXPOSE=1`) — see README "Using pi-studio".
 - **No Julia/R/GHCi/Clojure runtimes**. Use `uv run --with X` for
  Python REPLs; `apt install` other-language runtimes ad-hoc per
  container if needed.
@@ -32,6 +32,12 @@ Pre-v1.0.0 tags followed the pi npm version (`v{pi_version}[letter]`).
    gate **only** the studio tags, so a studio build/smoke failure can
    never block the core `:latest` / `:vX.Y.Z` release.
  - `STUDIO_PORT=8765` baked as an advisory default.
+- **`studio-expose` helper + `socat` (base).** Because pi-studio binds the
+  container's loopback, a published Docker port can't reach it. The new
+  `studio-expose` helper (socat, added to the base) bridges the container's
+  loopback to its egress interface on the same port; set `STUDIO_EXPOSE=1`
+  in compose to auto-start it on boot (default off — Studio stays
+  loopback-only otherwise). `socat` is in the base for all variants.
 - **README "Using pi-studio" section.** Documents the container access
  reality: pi-studio hard-binds `127.0.0.1` inside the container
  (`.listen(port,"127.0.0.1")`, no `--host` flag), so a plain `-p`
@@ -50,6 +50,10 @@ ENV DEBIAN_FRONTEND=noninteractive
 #   graphviz     — `dot` rendering for many diagram tools. ~10 MB.
 #   imagemagick  — image conversion / resizing for thumbnails, etc. ~50 MB.
 #   yq           — YAML-aware companion to jq.
+#   socat        — TCP relay. Powers `studio-expose`, which bridges
+#                  pi-studio's container-loopback server to the container's
+#                  external interface so a published port can reach it.
+#                  ~1 MB; generally useful for any port-forwarding need.
 RUN apt-get update && \
    apt-get upgrade -y --no-install-recommends && \
    apt-get install -y --no-install-recommends \
@@ -85,6 +89,7 @@ RUN apt-get update && \
    pandoc \
    graphviz \
    imagemagick \
+    socat \
    && ln -s /usr/bin/fdfind /usr/local/bin/fd \
    && apt-get clean \
    && rm -rf /var/lib/apt/lists/*
@@ -430,9 +435,11 @@ COPY rootfs/home/developer/.inputrc      /etc/skel-devbox/.inputrc

 # ── Entrypoint ────────────────────────────────────────────────────────
 COPY rootfs/usr/local/lib/pi-devbox/ /usr/local/lib/pi-devbox/
+COPY rootfs/usr/local/bin/studio-expose /usr/local/bin/studio-expose
 COPY entrypoint.sh /usr/local/bin/entrypoint.sh
 COPY entrypoint-user.sh /usr/local/bin/entrypoint-user.sh
 RUN chmod +x /usr/local/bin/entrypoint.sh /usr/local/bin/entrypoint-user.sh \
+    /usr/local/bin/studio-expose \
    /usr/local/lib/pi-devbox/*.sh 2>/dev/null || true

 # Start as root — entrypoint adjusts UID/GID then drops to developer
@@ -164,7 +164,7 @@ There is no `--host`/bind flag. This matters for a container: a plain
 interface, **not** its loopback, so it will not reach Studio. Two paths
 work:

-**A. Host networking (simplest — recommended on OrbStack / single-host).**
+**A. Host networking (simplest — OrbStack / single-host, no bridge).**
 Run the container with host networking so the container's loopback is the
 host's loopback:

@@ -175,28 +175,37 @@ services:
 ```

 Then `http://127.0.0.1:8765/?token=…` works in a browser on the Docker
-host. (Note: host networking changes `host.docker.internal` semantics, so
-weigh it against the LAN-jump SSH feature if you use that.)
+host. This is the most secure option (Studio never leaves loopback). Note:
+host networking changes `host.docker.internal` semantics, so weigh it
+against the LAN-jump SSH feature if you use that.

-**B. Loopback bridge (portable — bridge networking).** Publish a port and
-bridge the container's loopback to its external interface with a one-liner
-(uses the bundled `node`; binds the eth0 IP only, so it never clashes with
-Studio's own `127.0.0.1:8765` listener):
+**B. `studio-expose` bridge (portable — any networking mode).** Publish a
+port and run the bundled `studio-expose` helper, which uses `socat` to
+bridge the container's loopback to its external interface (binding the
+egress IP on the same port, so the token URL Studio printed works
+verbatim):

 ```yaml
 services:
  devbox:
    ports:
      - "127.0.0.1:8765:8765"   # host-localhost only
+    environment:
+      - STUDIO_EXPOSE=1          # auto-start the bridge on container boot
 ```

+With `STUDIO_EXPOSE=1`, the entrypoint starts the bridge for you; just run
+`/studio --port 8765` in your pi session. To bridge manually instead
+(leave `STUDIO_EXPOSE` unset), run `studio-expose` in a container shell:
+
 ```bash
-# inside the container, after /studio --port 8765:
-EXT=$(hostname -i)
-node -e 'const net=require("net"),p=process.env.STUDIO_PORT||8765,h=process.argv[1];\
-net.createServer(c=>{const u=net.connect(p,"127.0.0.1");c.pipe(u);u.pipe(c);u.on("error",()=>c.destroy());c.on("error",()=>u.destroy());}).listen(p,h,()=>console.log("bridge "+h+":"+p+" -> 127.0.0.1:"+p));' "$EXT"
+studio-expose            # bridges $STUDIO_PORT (default 8765); --help for details
 ```

+> **Security:** the bridge intentionally exposes Studio beyond loopback;
+> its tokenized URL is the only auth. Keep the host-side publish on
+> `127.0.0.1:` and use `ssh -L` for remote access. Default is **off**.
+
 ### Remote host (SSH / mosh)

 When the Docker host is remote, keep Studio on localhost and forward the
@@ -121,6 +121,25 @@ if command -v pi &>/dev/null; then
  done
 fi

+# ── pi-studio: optional loopback bridge (opt-in) ──────────────────────
+# pi-studio binds its server to 127.0.0.1 inside the container, which a
+# published Docker port cannot reach. When STUDIO_EXPOSE is truthy (set in
+# compose), start the `studio-expose` socat bridge in the background so a
+# published port + `ssh -L` tunnel can reach Studio once the user runs
+# `/studio --port "$STUDIO_PORT"`. Default OFF — Studio stays loopback-only
+# (its secure default) unless explicitly opted in. Guarded on the studio
+# variant (/opt/pi-studio) so it is a no-op in the plain image.
+case "${STUDIO_EXPOSE:-}" in
+  1|true|TRUE|yes|on)
+    if [ -d /opt/pi-studio ] && command -v studio-expose &>/dev/null && command -v socat &>/dev/null; then
+      echo "STUDIO_EXPOSE set — starting studio-expose bridge on port ${STUDIO_PORT:-8765} (background)"
+      nohup studio-expose "${STUDIO_PORT:-8765}" >/tmp/studio-expose.log 2>&1 &
+    else
+      echo "STUDIO_EXPOSE set but studio-expose/socat/pi-studio unavailable — skipping bridge"
+    fi
+    ;;
+esac
+
 # ── Skillset: deploy skills/instructions from mounted skillset repo ──
 # When the skillset repo is mounted (at $HOME/skillset or /workspace/skillset),
 # run the deploy script to create relative symlinks for skills and instructions.
@@ -0,0 +1,75 @@
+#!/usr/bin/env bash
+# studio-expose — make a container-loopback pi-studio server reachable
+# through a published Docker port.
+#
+# WHY THIS EXISTS
+#   pi-studio hard-binds its HTTP/WebSocket server to 127.0.0.1 inside the
+#   container (index.ts: `.listen(port, "127.0.0.1")`) and there is no
+#   --host / bind flag. A plain `docker run -p 8765:8765` forwards to the
+#   container's EXTERNAL interface (eth0), not its loopback, so it cannot
+#   reach Studio. This helper runs a socat TCP relay that listens on the
+#   container's egress IP and forwards to 127.0.0.1:<port>, so a published
+#   port (and an `ssh -L` tunnel from your laptop) can reach Studio.
+#
+# SECURITY
+#   This intentionally exposes Studio beyond loopback — anything that can
+#   reach the container's network interface (and the host port you publish)
+#   can connect. Studio's tokenized URL is the only auth. Mitigate by
+#   publishing the host port on localhost only:
+#       ports: ["127.0.0.1:${STUDIO_PORT}:${STUDIO_PORT}"]
+#   and use `ssh -L` for remote access. Bridge nothing you don't intend to.
+#
+# USAGE
+#   studio-expose [PORT]        # bridge PORT (default: $STUDIO_PORT or 8765)
+#   studio-expose --help
+#
+#   Typically: inside a pi session run `/studio --no-browser --port 8765`,
+#   then in a container shell run `studio-expose` (or set STUDIO_EXPOSE=1 in
+#   compose to auto-start it on container boot — see entrypoint-user.sh).
+#
+# Runs in the foreground; Ctrl-C to stop. The entrypoint auto-start path
+# runs it backgrounded.
+set -euo pipefail
+
+PORT="${1:-${STUDIO_PORT:-8765}}"
+
+if [ "$PORT" = "--help" ] || [ "$PORT" = "-h" ]; then
+  sed -n '2,31p' "$0" | sed 's/^# \{0,1\}//'
+  exit 0
+fi
+
+case "$PORT" in
+  ''|*[!0-9]*) echo "studio-expose: invalid port '$PORT'" >&2; exit 2 ;;
+esac
+
+if ! command -v socat >/dev/null 2>&1; then
+  echo "studio-expose: socat not found in PATH" >&2
+  exit 1
+fi
+
+# Container's primary egress IPv4. In Docker the container hostname resolves
+# to its eth0 address, so `hostname -I` lists it; we take the first
+# non-loopback IPv4. We must bind this specific address rather than 0.0.0.0
+# — binding 0.0.0.0 would collide with Studio's own 127.0.0.1:PORT listener
+# (0.0.0.0 includes loopback) and fail with EADDRINUSE. `ip route get` is a
+# fallback only when iproute2 happens to be present (not in the base image).
+BIND_IP="$(hostname -I 2>/dev/null | tr ' ' '\n' \
+  | grep -E '^[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+$' | grep -vE '^127\.' | head -n1)"
+if [ -z "${BIND_IP:-}" ] && command -v ip >/dev/null 2>&1; then
+  BIND_IP="$(ip -4 route get 1.1.1.1 2>/dev/null | awk '{for(i=1;i<=NF;i++) if($i=="src"){print $(i+1); exit}}')"
+fi
+[ -n "${BIND_IP:-}" ] || BIND_IP="$(hostname -i 2>/dev/null | awk '{print $1}')"
+if [ -z "${BIND_IP:-}" ]; then
+  echo "studio-expose: could not determine container egress IP" >&2
+  exit 1
+fi
+
+echo "studio-expose: bridging ${BIND_IP}:${PORT} -> 127.0.0.1:${PORT}"
+echo "studio-expose: open the tokenized URL pi-studio printed; if the host"
+echo "studio-expose: publishes ${PORT}, reach it at http://127.0.0.1:${PORT}/?token=..."
+echo "studio-expose: (remote host: ssh -L ${PORT}:127.0.0.1:${PORT} user@host)"
+
+# fork: one child per connection (handles concurrent + long-lived WebSocket
+# connections). reuseaddr: survive quick restarts. Studio need not be up yet
+# — connections simply fail until `/studio --port ${PORT}` is running.
+exec socat "TCP-LISTEN:${PORT},bind=${BIND_IP},fork,reuseaddr" "TCP:127.0.0.1:${PORT}"
@@ -78,6 +78,8 @@ run "graphviz (dot)"  "dot -V"
 run "imagemagick"     "magick --version"
 run "yq"              "yq --version"
 run "tldr (tealdeer)" "tldr --version"
+run "socat"           "socat -V"
+run "studio-expose helper" "test -x /usr/local/bin/studio-expose"

 # ── tmux 0-indexing (required for pi-studio variants) ─────────────────
 echo ""