Files
molecule-ai-workspace-templ…/Dockerfile
Molecule AI Dev Engineer A (Kimi) 74493abdc6
CI / Template validation (static) (push) Successful in 3s
CI / Adapter unit tests (push) Successful in 7s
Secret scan / Scan diff for credential-shaped strings (pull_request) Successful in 11s
CI / Template validation (static) (pull_request) Successful in 13s
CI / Adapter unit tests (pull_request) Successful in 14s
verify-providers-projection / Regenerate projection, fail on drift, assert registry ⊆ template (pull_request) Failing after 1m9s
CI / Template validation (runtime) (push) Successful in 1m42s
CI / Template validation (runtime) (pull_request) Successful in 1m9s
CI / T4 tier-4 conformance (live) (push) Successful in 3m44s
CI / T4 tier-4 conformance (live) (pull_request) Successful in 3m11s
CI / validate (pull_request) Successful in 8s
CI / validate (push) Successful in 9s
fix(dockerfile): fail-closed claude install + PATH verification (#75)
- Remove `2>/dev/null || true` masking from `npm install -g @anthropic-ai/claude-code`
  so install failures fail the image build instead of silently producing a
  green image without the primary runtime engine.
- Add `command -v claude` PATH assertion to catch prefix misconfigs,
  package renames, or masked install failures.

Split from PR #79; CI aggregate changes moved to separate PR.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
2026-06-03 18:46:43 +00:00

173 lines
9.0 KiB
Docker

FROM python:3.11-slim
# System deps — curl/gosu/node/npm for the runtime; git for agent
# autonomy against the Molecule-owned Gitea middleman.
# Without these the team's claim-and-ship loop silently returns
# "(no response generated)" because tools error out.
#
# T4 escalation leg (RFC internal#456 §9 / PR#474):
# sudo + util-linux(nsenter) + docker.io(CLI) are baked here so the
# uid-1000 `agent` (see useradd below — UNCHANGED, agent stays
# uid-1000) has a wired, audited path to host root inside the
# provisioner's `--privileged --pid=host -v /:/host
# -v /var/run/docker.sock:/var/run/docker.sock` container. Without
# sudo, a uid-1000 process in --privileged CANNOT nsenter/chroot
# /host (--privileged grants caps to root, not uid-1000) and cannot
# use the root:docker 0660 docker.sock — T4 would be
# provisioner-shape-only (the documented ABSENT-escalation-leg gap).
# The sudoers drop-in + docker-group add are below, after useradd,
# so `agent` exists. This is ADDITIVE: it does NOT change the agent
# uid and does NOT change /configs token ownership (still uid-1000,
# enforced by entrypoint.sh + the Layer-3 conformance gate).
#
# rsync (added cp#326 2026-05-26):
# entrypoint.sh's restore_from_secondary_volume() rsyncs the prior
# workspace's /configs, /workspace, and /home/agent/.claude back
# into root on first boot from the snapshot-restored secondary
# volume CP attaches at /dev/xvdb. Without rsync the restore path
# silently no-ops and the workspace boots with empty state.
#
# e2fsprogs (added cp#326 2026-05-26):
# provides /sbin/blkid + /sbin/e2label so the restore code can
# probe the secondary volume's filesystem type before mounting.
# Already pulled in indirectly by util-linux on Debian-slim but
# pinning it explicit makes the dependency self-documenting and
# survives a future base-image change.
RUN apt-get update && apt-get install -y --no-install-recommends \
curl gosu nodejs npm ca-certificates git sudo util-linux docker.io xdotool scrot \
rsync e2fsprogs \
&& rm -rf /var/lib/apt/lists/*
# Install claude-code CLI via npm — fail-closed so an npm outage, package
# rename, auth failure, or Node/npm breakage fails the image build instead
# of producing a green image without the primary runtime engine (#75).
RUN npm install -g @anthropic-ai/claude-code
# Verify the CLI resolved in PATH (catches masked install failures,
# package renames, or npm prefix misconfig).
RUN command -v claude >/dev/null 2>&1 || (echo "ERROR: claude CLI not found in PATH" >&2 && exit 1)
# Create agent user — UNCHANGED. The agent runs as uid-1000; the T4
# escalation leg below is additive and does NOT promote the agent to
# root. claude-code still refuses --dangerously-skip-permissions as
# root, and /configs/.auth_token must stay agent-owned (Hermes
# list_peers 401 class — RFC internal#456 §10).
RUN useradd -u 1000 -m -s /bin/bash agent && \
mkdir -p /agent-home && \
chown agent:agent /agent-home
# --- T4 escalation leg (RFC internal#456 §9.3 / PR#474) ---
# Wired path: uid-1000 agent -> host root inside the provisioner's
# --privileged --pid=host -v /:/host -v docker.sock container.
# 1. NOPASSWD sudoers drop-in (mode 0440, visudo-validated at build
# so a malformed sudoers can never ship a broken-sudo image).
# 2. agent in the `docker` group so the bind-mounted root:docker
# 0660 /var/run/docker.sock is usable without sudo.
# Atomic co-sequencing (RFC §10): this ships in the SAME image
# revision as the uid-1000 + agent-owned-token entrypoint contract;
# the Layer-3 conformance gate asserts BOTH on the running container.
RUN set -eux; \
printf 'agent ALL=(ALL) NOPASSWD:ALL\n' > /etc/sudoers.d/agent-t4; \
chmod 0440 /etc/sudoers.d/agent-t4; \
visudo -cf /etc/sudoers.d/agent-t4; \
groupadd -f docker; \
groupadd -g 988 -f docker-host || true; \
usermod -aG docker agent; \
usermod -aG docker-host agent || true; \
id agent
WORKDIR /app
# RUNTIME_VERSION is forwarded from the reusable publish workflow as
# a docker build-arg. When set (cascade-triggered builds), it's the
# exact runtime version PyPI just published. Including it as an ARG
# changes the cache key for the pip install layer below — without
# this, identical Dockerfile + identical requirements.txt content
# would let docker reuse the cached layer with the previous version
# baked in (the cache trap that bit us 5x on 2026-04-27).
# Empty default = falls back to whatever requirements.txt resolves to.
ARG RUNTIME_VERSION=
ARG PIP_INDEX_URL=https://git.moleculesai.app/api/packages/molecule-ai/pypi/simple/
# Install Python deps. The RUNTIME_VERSION ARG is a no-op argument to
# the RUN command itself but its presence as a declared ARG above
# means buildx hashes it into the cache key.
COPY requirements.txt .
RUN pip install --no-cache-dir --index-url "${PIP_INDEX_URL}" -r requirements.txt && \
if [ -n "${RUNTIME_VERSION}" ]; then \
pip install --no-cache-dir --index-url "${PIP_INDEX_URL}" --upgrade "molecule-ai-workspace-runtime==${RUNTIME_VERSION}"; \
fi
# MOLECULE-HOTFIX (claude-code 2.1.150 / agent-sdk 0.2.84): apply in-place
# SDK patch so the receive_messages loop treats is_error+subtype=success as
# end-of-stream rather than raising Exception("... success"). See
# scripts/patch_claude_sdk_2_1_150.py for the rationale + upstream removal
# criteria. Idempotent; fails the build if the upstream SDK has been
# updated so we notice the workaround is stale.
COPY scripts/patch_claude_sdk_2_1_150.py /tmp/patch_claude_sdk_2_1_150.py
RUN python3 /tmp/patch_claude_sdk_2_1_150.py && rm /tmp/patch_claude_sdk_2_1_150.py
# Copy adapter code
COPY adapter.py .
COPY __init__.py .
# Provider registry. The adapter's _load_providers walks 4 paths:
# 1. /opt/adapter/config.yaml — provisioner-managed canonical
# 2. os.path.dirname(__file__)/config.yaml — alongside adapter.py (this image)
# 3. ${WORKSPACE_CONFIG_PATH}/config.yaml — workspace per-instance overrides
# 4. _BUILTIN_PROVIDERS — oauth + anthropic-api only
# On this image /opt/adapter/ is never populated by the platform
# provisioner, so path 2 (/app/config.yaml) is the load-bearing one.
# Without this COPY the file isn't in the image, all 3 file paths fail,
# and _load_providers falls through to _BUILTIN_PROVIDERS — every
# MiniMax/GLM/Kimi/DeepSeek model silently routes to anthropic-oauth →
# "Not logged in. Please run /login" at first LLM call. Caused the
# canary's 38h chronic red on 2026-05-07/08 (molecule-core#129).
COPY config.yaml .
# Adapter-specific executor — owned by THIS template (universal-runtime
# refactor, molecule-core task #87). Lives alongside adapter.py so
# Python's import system picks the local /app/claude_sdk_executor.py
# before the same-named module that older molecule-runtime versions
# also shipped under site-packages. Once molecule-core drops the file
# from its workspace/ package and bumps the runtime PyPI version, the
# template will be the sole source of truth.
COPY claude_sdk_executor.py .
# Set the adapter module for runtime discovery
ENV ADAPTER_MODULE=adapter
# Optional GitHub mirror credential helper + background refresh daemon.
# GitHub is not on the critical path; these scripts are inert unless
# ENABLE_GITHUB_MIRROR_CREDENTIALS=true is set by an operator.
COPY scripts/molecule-git-token-helper.sh /app/scripts/molecule-git-token-helper.sh
COPY scripts/molecule-gh-token-refresh.sh /app/scripts/molecule-gh-token-refresh.sh
RUN chmod +x /app/scripts/molecule-git-token-helper.sh /app/scripts/molecule-gh-token-refresh.sh
# Generic GIT_ASKPASS helper — image-side companion to molecule-core PR
# #1525 (workspace-server applyAgentGitIdentity, merge_sha 73a09443a086).
# Reads HTTPS Basic-Auth credentials from env vars (GIT_HTTP_USERNAME /
# GIT_HTTP_PASSWORD, with GITEA_USER / GITEA_TOKEN as fallback) and emits
# them on the git credential-prompt protocol, so container-side `git` can
# authenticate to any private HTTPS remote without on-disk ~/.gitconfig
# or ~/.git-credentials mutation. The platform provisioner sets
# GIT_ASKPASS=/usr/local/bin/molecule-askpass via applyAgentGitIdentity;
# until this binary ships in the runtime image, git invocations error
# with "exec: /usr/local/bin/molecule-askpass: not found" (forward-only
# pin gap — same class as Hermes list_peers and codex template breakage,
# fixed image-side here).
#
# No hardcoded hostnames or vendor names — the script body is identical
# to the one shipped in molecule-core workspace/scripts/molecule-askpass
# and the parallel external workspace template repos, so any deployer
# can fork this template and use it against their own git host without
# editing.
COPY scripts/molecule-askpass /usr/local/bin/molecule-askpass
RUN chmod +x /usr/local/bin/molecule-askpass
# Drop-priv entrypoint — claude-code refuses --dangerously-skip-permissions
# as root, so we run molecule-runtime as the agent user (uid 1000).
# The script handles volume-ownership fix + session-dir symlink before
# exec'ing via gosu.
COPY entrypoint.sh /entrypoint.sh
RUN chmod +x /entrypoint.sh
ENTRYPOINT ["/entrypoint.sh"]