Fix/shadow env in container (#646)

* fix: shadow .env file in container to prevent agents from reading secrets

The main agent's container mounts the project root read-only, which
inadvertently exposed the .env file containing API keys. Mount /dev/null
over /workspace/project/.env to shadow it — secrets are already passed
via stdin and never need to be read from disk inside the container.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: adapt .env shadowing and runtime for Apple Container

Apple Container (VirtioFS) only supports directory mounts, not file
mounts. The previous /dev/null host-side mount over .env crashes with
VZErrorDomain "A directory sharing device configuration is invalid".

- Dockerfile: entrypoint now shadows .env via mount --bind inside the
  container, then drops privileges via setpriv to the host UID/GID
- container-runner: main containers skip --user and pass RUN_UID/RUN_GID
  env vars so entrypoint starts as root for mount --bind
- container-runtime: switch to Apple Container CLI (container), fix
  cleanupOrphans to use container list --format json
- Skill: add Dockerfile and container-runner.ts to
  convert-to-apple-container skill (v1.1.0)

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

* fix: revert src to Docker runtime, keep Apple Container in skill only

The source files should remain Docker-compatible. The Apple Container
adaptations live in the convert-to-apple-container skill and are applied
on demand.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
gavrielc
2026-03-02 13:28:28 +02:00
committed by GitHub
parent 770231687a
commit bae8538695
7 changed files with 850 additions and 3 deletions

View File

@@ -13,11 +13,13 @@ This skill switches NanoClaw's container runtime from Docker to Apple Container
- Startup check: `docker info``container system status` (with auto-start)
- Orphan detection: `docker ps --filter``container ls --format json`
- Build script default: `docker``container`
- Dockerfile entrypoint: `.env` shadowing via `mount --bind` inside the container (Apple Container only supports directory mounts, not file mounts like Docker's `/dev/null` overlay)
- Container runner: main-group containers start as root for `mount --bind`, then drop privileges via `setpriv`
**What stays the same:**
- Dockerfile (shared by both runtimes)
- Container runner code (`src/container-runner.ts`)
- Mount security/allowlist validation
- All exported interfaces and IPC protocol
- Non-main container behavior (still uses `--user` flag)
- All other functionality
## Prerequisites
@@ -72,11 +74,15 @@ npx tsx scripts/apply-skill.ts .claude/skills/convert-to-apple-container
This deterministically:
- Replaces `src/container-runtime.ts` with the Apple Container implementation
- Replaces `src/container-runtime.test.ts` with Apple Container-specific tests
- Updates `src/container-runner.ts` with .env shadow mount fix and privilege dropping
- Updates `container/Dockerfile` with entrypoint that shadows .env via `mount --bind`
- Updates `container/build.sh` to default to `container` runtime
- Records the application in `.nanoclaw/state.yaml`
If the apply reports merge conflicts, read the intent files:
- `modify/src/container-runtime.ts.intent.md` — what changed and invariants
- `modify/src/container-runner.ts.intent.md` — .env shadow and privilege drop changes
- `modify/container/Dockerfile.intent.md` — entrypoint changes for .env shadowing
- `modify/container/build.sh.intent.md` — what changed for build script
### Validate code changes
@@ -172,4 +178,6 @@ Check directory permissions on the host. The container runs as uid 1000.
|------|----------------|
| `src/container-runtime.ts` | Full replacement — Docker → Apple Container API |
| `src/container-runtime.test.ts` | Full replacement — tests for Apple Container behavior |
| `src/container-runner.ts` | .env shadow mount removed, main containers start as root with privilege drop |
| `container/Dockerfile` | Entrypoint: `mount --bind` for .env shadowing, `setpriv` privilege drop |
| `container/build.sh` | Default runtime: `docker``container` |

View File

@@ -1,12 +1,14 @@
skill: convert-to-apple-container
version: 1.0.0
version: 1.1.0
description: "Switch container runtime from Docker to Apple Container (macOS)"
core_version: 0.1.0
adds: []
modifies:
- src/container-runtime.ts
- src/container-runtime.test.ts
- src/container-runner.ts
- container/build.sh
- container/Dockerfile
structured: {}
conflicts: []
depends: []

View File

@@ -0,0 +1,68 @@
# NanoClaw Agent Container
# Runs Claude Agent SDK in isolated Linux VM with browser automation
FROM node:22-slim
# Install system dependencies for Chromium
RUN apt-get update && apt-get install -y \
chromium \
fonts-liberation \
fonts-noto-color-emoji \
libgbm1 \
libnss3 \
libatk-bridge2.0-0 \
libgtk-3-0 \
libx11-xcb1 \
libxcomposite1 \
libxdamage1 \
libxrandr2 \
libasound2 \
libpangocairo-1.0-0 \
libcups2 \
libdrm2 \
libxshmfence1 \
curl \
git \
&& rm -rf /var/lib/apt/lists/*
# Set Chromium path for agent-browser
ENV AGENT_BROWSER_EXECUTABLE_PATH=/usr/bin/chromium
ENV PLAYWRIGHT_CHROMIUM_EXECUTABLE_PATH=/usr/bin/chromium
# Install agent-browser and claude-code globally
RUN npm install -g agent-browser @anthropic-ai/claude-code
# Create app directory
WORKDIR /app
# Copy package files first for better caching
COPY agent-runner/package*.json ./
# Install dependencies
RUN npm install
# Copy source code
COPY agent-runner/ ./
# Build TypeScript
RUN npm run build
# Create workspace directories
RUN mkdir -p /workspace/group /workspace/global /workspace/extra /workspace/ipc/messages /workspace/ipc/tasks /workspace/ipc/input
# Create entrypoint script
# Secrets are passed via stdin JSON — temp file is deleted immediately after Node reads it
# Follow-up messages arrive via IPC files in /workspace/ipc/input/
# Apple Container only supports directory mounts (VirtioFS), so .env cannot be
# shadowed with a host-side /dev/null file mount. Instead the entrypoint starts
# as root, uses mount --bind to shadow .env, then drops to the host user via setpriv.
RUN printf '#!/bin/bash\nset -e\n\n# Shadow .env so the agent cannot read host secrets (requires root)\nif [ "$(id -u)" = "0" ] && [ -f /workspace/project/.env ]; then\n mount --bind /dev/null /workspace/project/.env\nfi\n\n# Compile agent-runner\ncd /app && npx tsc --outDir /tmp/dist 2>&1 >&2\nln -s /app/node_modules /tmp/dist/node_modules\nchmod -R a-w /tmp/dist\n\n# Capture stdin (secrets JSON) to temp file\ncat > /tmp/input.json\n\n# Drop privileges if running as root (main-group containers)\nif [ "$(id -u)" = "0" ] && [ -n "$RUN_UID" ]; then\n chown "$RUN_UID:$RUN_GID" /tmp/input.json /tmp/dist\n exec setpriv --reuid="$RUN_UID" --regid="$RUN_GID" --clear-groups -- node /tmp/dist/index.js < /tmp/input.json\nfi\n\nexec node /tmp/dist/index.js < /tmp/input.json\n' > /app/entrypoint.sh && chmod +x /app/entrypoint.sh
# Set ownership to node user (non-root) for writable directories
RUN chown -R node:node /workspace && chmod 777 /home/node
# Set working directory to group workspace
WORKDIR /workspace/group
# Entry point reads JSON from stdin, outputs JSON to stdout
ENTRYPOINT ["/app/entrypoint.sh"]

View File

@@ -0,0 +1,31 @@
# Intent: container/Dockerfile modifications
## What changed
Updated the entrypoint script to shadow `.env` inside the container and drop privileges at runtime, replacing the Docker-style host-side file mount approach.
## Why
Apple Container (VirtioFS) only supports directory mounts, not file mounts. The Docker approach of mounting `/dev/null` over `.env` from the host causes `VZErrorDomain Code=2 "A directory sharing device configuration is invalid"`. The fix moves the shadowing into the entrypoint using `mount --bind` (which works inside the Linux VM).
## Key sections
### Entrypoint script
- Added: `mount --bind /dev/null /workspace/project/.env` when running as root and `.env` exists
- Added: Privilege drop via `setpriv --reuid=$RUN_UID --regid=$RUN_GID --clear-groups` for main-group containers
- Added: `chown` of `/tmp/input.json` and `/tmp/dist` to target user before dropping privileges
- Removed: `USER node` directive — main containers start as root to perform the bind mount, then drop privileges in the entrypoint. Non-main containers still get `--user` from the host.
### Dual-path execution
- Root path (main containers): shadow .env → compile → capture stdin → chown → setpriv drop → exec node
- Non-root path (other containers): compile → capture stdin → exec node
## Invariants
- The entrypoint still reads JSON from stdin and runs the agent-runner
- The compiled output goes to `/tmp/dist` (read-only after build)
- `node_modules` is symlinked, not copied
- Non-main containers are unaffected (they arrive as non-root via `--user`)
## Must-keep
- The `set -e` at the top
- The stdin capture to `/tmp/input.json` (required because setpriv can't forward stdin piping)
- The `chmod -R a-w /tmp/dist` (prevents agent from modifying its own runner)
- The `chown -R node:node /workspace` in the build step

View File

@@ -0,0 +1,694 @@
/**
* Container Runner for NanoClaw
* Spawns agent execution in containers and handles IPC
*/
import { ChildProcess, exec, spawn } from 'child_process';
import fs from 'fs';
import path from 'path';
import {
CONTAINER_IMAGE,
CONTAINER_MAX_OUTPUT_SIZE,
CONTAINER_TIMEOUT,
DATA_DIR,
GROUPS_DIR,
IDLE_TIMEOUT,
TIMEZONE,
} from './config.js';
import { readEnvFile } from './env.js';
import { resolveGroupFolderPath, resolveGroupIpcPath } from './group-folder.js';
import { logger } from './logger.js';
import {
CONTAINER_RUNTIME_BIN,
readonlyMountArgs,
stopContainer,
} from './container-runtime.js';
import { validateAdditionalMounts } from './mount-security.js';
import { RegisteredGroup } from './types.js';
// Sentinel markers for robust output parsing (must match agent-runner)
const OUTPUT_START_MARKER = '---NANOCLAW_OUTPUT_START---';
const OUTPUT_END_MARKER = '---NANOCLAW_OUTPUT_END---';
export interface ContainerInput {
prompt: string;
sessionId?: string;
groupFolder: string;
chatJid: string;
isMain: boolean;
isScheduledTask?: boolean;
assistantName?: string;
secrets?: Record<string, string>;
}
export interface ContainerOutput {
status: 'success' | 'error';
result: string | null;
newSessionId?: string;
error?: string;
}
interface VolumeMount {
hostPath: string;
containerPath: string;
readonly: boolean;
}
function buildVolumeMounts(
group: RegisteredGroup,
isMain: boolean,
): VolumeMount[] {
const mounts: VolumeMount[] = [];
const projectRoot = process.cwd();
const groupDir = resolveGroupFolderPath(group.folder);
if (isMain) {
// Main gets the project root read-only. Writable paths the agent needs
// (group folder, IPC, .claude/) are mounted separately below.
// Read-only prevents the agent from modifying host application code
// (src/, dist/, package.json, etc.) which would bypass the sandbox
// entirely on next restart.
mounts.push({
hostPath: projectRoot,
containerPath: '/workspace/project',
readonly: true,
});
// Main also gets its group folder as the working directory
mounts.push({
hostPath: groupDir,
containerPath: '/workspace/group',
readonly: false,
});
} else {
// Other groups only get their own folder
mounts.push({
hostPath: groupDir,
containerPath: '/workspace/group',
readonly: false,
});
// Global memory directory (read-only for non-main)
// Only directory mounts are supported, not file mounts
const globalDir = path.join(GROUPS_DIR, 'global');
if (fs.existsSync(globalDir)) {
mounts.push({
hostPath: globalDir,
containerPath: '/workspace/global',
readonly: true,
});
}
}
// Per-group Claude sessions directory (isolated from other groups)
// Each group gets their own .claude/ to prevent cross-group session access
const groupSessionsDir = path.join(
DATA_DIR,
'sessions',
group.folder,
'.claude',
);
fs.mkdirSync(groupSessionsDir, { recursive: true });
const settingsFile = path.join(groupSessionsDir, 'settings.json');
if (!fs.existsSync(settingsFile)) {
fs.writeFileSync(
settingsFile,
JSON.stringify(
{
env: {
// Enable agent swarms (subagent orchestration)
// https://code.claude.com/docs/en/agent-teams#orchestrate-teams-of-claude-code-sessions
CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS: '1',
// Load CLAUDE.md from additional mounted directories
// https://code.claude.com/docs/en/memory#load-memory-from-additional-directories
CLAUDE_CODE_ADDITIONAL_DIRECTORIES_CLAUDE_MD: '1',
// Enable Claude's memory feature (persists user preferences between sessions)
// https://code.claude.com/docs/en/memory#manage-auto-memory
CLAUDE_CODE_DISABLE_AUTO_MEMORY: '0',
},
},
null,
2,
) + '\n',
);
}
// Sync skills from container/skills/ into each group's .claude/skills/
const skillsSrc = path.join(process.cwd(), 'container', 'skills');
const skillsDst = path.join(groupSessionsDir, 'skills');
if (fs.existsSync(skillsSrc)) {
for (const skillDir of fs.readdirSync(skillsSrc)) {
const srcDir = path.join(skillsSrc, skillDir);
if (!fs.statSync(srcDir).isDirectory()) continue;
const dstDir = path.join(skillsDst, skillDir);
fs.cpSync(srcDir, dstDir, { recursive: true });
}
}
mounts.push({
hostPath: groupSessionsDir,
containerPath: '/home/node/.claude',
readonly: false,
});
// Per-group IPC namespace: each group gets its own IPC directory
// This prevents cross-group privilege escalation via IPC
const groupIpcDir = resolveGroupIpcPath(group.folder);
fs.mkdirSync(path.join(groupIpcDir, 'messages'), { recursive: true });
fs.mkdirSync(path.join(groupIpcDir, 'tasks'), { recursive: true });
fs.mkdirSync(path.join(groupIpcDir, 'input'), { recursive: true });
mounts.push({
hostPath: groupIpcDir,
containerPath: '/workspace/ipc',
readonly: false,
});
// Copy agent-runner source into a per-group writable location so agents
// can customize it (add tools, change behavior) without affecting other
// groups. Recompiled on container startup via entrypoint.sh.
const agentRunnerSrc = path.join(
projectRoot,
'container',
'agent-runner',
'src',
);
const groupAgentRunnerDir = path.join(
DATA_DIR,
'sessions',
group.folder,
'agent-runner-src',
);
if (!fs.existsSync(groupAgentRunnerDir) && fs.existsSync(agentRunnerSrc)) {
fs.cpSync(agentRunnerSrc, groupAgentRunnerDir, { recursive: true });
}
mounts.push({
hostPath: groupAgentRunnerDir,
containerPath: '/app/src',
readonly: false,
});
// Additional mounts validated against external allowlist (tamper-proof from containers)
if (group.containerConfig?.additionalMounts) {
const validatedMounts = validateAdditionalMounts(
group.containerConfig.additionalMounts,
group.name,
isMain,
);
mounts.push(...validatedMounts);
}
return mounts;
}
/**
* Read allowed secrets from .env for passing to the container via stdin.
* Secrets are never written to disk or mounted as files.
*/
function readSecrets(): Record<string, string> {
return readEnvFile(['CLAUDE_CODE_OAUTH_TOKEN', 'ANTHROPIC_API_KEY']);
}
function buildContainerArgs(
mounts: VolumeMount[],
containerName: string,
isMain: boolean,
): string[] {
const args: string[] = ['run', '-i', '--rm', '--name', containerName];
// Pass host timezone so container's local time matches the user's
args.push('-e', `TZ=${TIMEZONE}`);
// Run as host user so bind-mounted files are accessible.
// Skip when running as root (uid 0), as the container's node user (uid 1000),
// or when getuid is unavailable (native Windows without WSL).
const hostUid = process.getuid?.();
const hostGid = process.getgid?.();
if (hostUid != null && hostUid !== 0 && hostUid !== 1000) {
if (isMain) {
// Main containers start as root so the entrypoint can mount --bind
// to shadow .env. Privileges are dropped via setpriv in entrypoint.sh.
args.push('-e', `RUN_UID=${hostUid}`);
args.push('-e', `RUN_GID=${hostGid}`);
} else {
args.push('--user', `${hostUid}:${hostGid}`);
}
args.push('-e', 'HOME=/home/node');
}
for (const mount of mounts) {
if (mount.readonly) {
args.push(...readonlyMountArgs(mount.hostPath, mount.containerPath));
} else {
args.push('-v', `${mount.hostPath}:${mount.containerPath}`);
}
}
args.push(CONTAINER_IMAGE);
return args;
}
export async function runContainerAgent(
group: RegisteredGroup,
input: ContainerInput,
onProcess: (proc: ChildProcess, containerName: string) => void,
onOutput?: (output: ContainerOutput) => Promise<void>,
): Promise<ContainerOutput> {
const startTime = Date.now();
const groupDir = resolveGroupFolderPath(group.folder);
fs.mkdirSync(groupDir, { recursive: true });
const mounts = buildVolumeMounts(group, input.isMain);
const safeName = group.folder.replace(/[^a-zA-Z0-9-]/g, '-');
const containerName = `nanoclaw-${safeName}-${Date.now()}`;
const containerArgs = buildContainerArgs(mounts, containerName, input.isMain);
logger.debug(
{
group: group.name,
containerName,
mounts: mounts.map(
(m) =>
`${m.hostPath} -> ${m.containerPath}${m.readonly ? ' (ro)' : ''}`,
),
containerArgs: containerArgs.join(' '),
},
'Container mount configuration',
);
logger.info(
{
group: group.name,
containerName,
mountCount: mounts.length,
isMain: input.isMain,
},
'Spawning container agent',
);
const logsDir = path.join(groupDir, 'logs');
fs.mkdirSync(logsDir, { recursive: true });
return new Promise((resolve) => {
const container = spawn(CONTAINER_RUNTIME_BIN, containerArgs, {
stdio: ['pipe', 'pipe', 'pipe'],
});
onProcess(container, containerName);
let stdout = '';
let stderr = '';
let stdoutTruncated = false;
let stderrTruncated = false;
// Pass secrets via stdin (never written to disk or mounted as files)
input.secrets = readSecrets();
container.stdin.write(JSON.stringify(input));
container.stdin.end();
// Remove secrets from input so they don't appear in logs
delete input.secrets;
// Streaming output: parse OUTPUT_START/END marker pairs as they arrive
let parseBuffer = '';
let newSessionId: string | undefined;
let outputChain = Promise.resolve();
container.stdout.on('data', (data) => {
const chunk = data.toString();
// Always accumulate for logging
if (!stdoutTruncated) {
const remaining = CONTAINER_MAX_OUTPUT_SIZE - stdout.length;
if (chunk.length > remaining) {
stdout += chunk.slice(0, remaining);
stdoutTruncated = true;
logger.warn(
{ group: group.name, size: stdout.length },
'Container stdout truncated due to size limit',
);
} else {
stdout += chunk;
}
}
// Stream-parse for output markers
if (onOutput) {
parseBuffer += chunk;
let startIdx: number;
while ((startIdx = parseBuffer.indexOf(OUTPUT_START_MARKER)) !== -1) {
const endIdx = parseBuffer.indexOf(OUTPUT_END_MARKER, startIdx);
if (endIdx === -1) break; // Incomplete pair, wait for more data
const jsonStr = parseBuffer
.slice(startIdx + OUTPUT_START_MARKER.length, endIdx)
.trim();
parseBuffer = parseBuffer.slice(endIdx + OUTPUT_END_MARKER.length);
try {
const parsed: ContainerOutput = JSON.parse(jsonStr);
if (parsed.newSessionId) {
newSessionId = parsed.newSessionId;
}
hadStreamingOutput = true;
// Activity detected — reset the hard timeout
resetTimeout();
// Call onOutput for all markers (including null results)
// so idle timers start even for "silent" query completions.
outputChain = outputChain.then(() => onOutput(parsed));
} catch (err) {
logger.warn(
{ group: group.name, error: err },
'Failed to parse streamed output chunk',
);
}
}
}
});
container.stderr.on('data', (data) => {
const chunk = data.toString();
const lines = chunk.trim().split('\n');
for (const line of lines) {
if (line) logger.debug({ container: group.folder }, line);
}
// Don't reset timeout on stderr — SDK writes debug logs continuously.
// Timeout only resets on actual output (OUTPUT_MARKER in stdout).
if (stderrTruncated) return;
const remaining = CONTAINER_MAX_OUTPUT_SIZE - stderr.length;
if (chunk.length > remaining) {
stderr += chunk.slice(0, remaining);
stderrTruncated = true;
logger.warn(
{ group: group.name, size: stderr.length },
'Container stderr truncated due to size limit',
);
} else {
stderr += chunk;
}
});
let timedOut = false;
let hadStreamingOutput = false;
const configTimeout = group.containerConfig?.timeout || CONTAINER_TIMEOUT;
// Grace period: hard timeout must be at least IDLE_TIMEOUT + 30s so the
// graceful _close sentinel has time to trigger before the hard kill fires.
const timeoutMs = Math.max(configTimeout, IDLE_TIMEOUT + 30_000);
const killOnTimeout = () => {
timedOut = true;
logger.error(
{ group: group.name, containerName },
'Container timeout, stopping gracefully',
);
exec(stopContainer(containerName), { timeout: 15000 }, (err) => {
if (err) {
logger.warn(
{ group: group.name, containerName, err },
'Graceful stop failed, force killing',
);
container.kill('SIGKILL');
}
});
};
let timeout = setTimeout(killOnTimeout, timeoutMs);
// Reset the timeout whenever there's activity (streaming output)
const resetTimeout = () => {
clearTimeout(timeout);
timeout = setTimeout(killOnTimeout, timeoutMs);
};
container.on('close', (code) => {
clearTimeout(timeout);
const duration = Date.now() - startTime;
if (timedOut) {
const ts = new Date().toISOString().replace(/[:.]/g, '-');
const timeoutLog = path.join(logsDir, `container-${ts}.log`);
fs.writeFileSync(
timeoutLog,
[
`=== Container Run Log (TIMEOUT) ===`,
`Timestamp: ${new Date().toISOString()}`,
`Group: ${group.name}`,
`Container: ${containerName}`,
`Duration: ${duration}ms`,
`Exit Code: ${code}`,
`Had Streaming Output: ${hadStreamingOutput}`,
].join('\n'),
);
// Timeout after output = idle cleanup, not failure.
// The agent already sent its response; this is just the
// container being reaped after the idle period expired.
if (hadStreamingOutput) {
logger.info(
{ group: group.name, containerName, duration, code },
'Container timed out after output (idle cleanup)',
);
outputChain.then(() => {
resolve({
status: 'success',
result: null,
newSessionId,
});
});
return;
}
logger.error(
{ group: group.name, containerName, duration, code },
'Container timed out with no output',
);
resolve({
status: 'error',
result: null,
error: `Container timed out after ${configTimeout}ms`,
});
return;
}
const timestamp = new Date().toISOString().replace(/[:.]/g, '-');
const logFile = path.join(logsDir, `container-${timestamp}.log`);
const isVerbose =
process.env.LOG_LEVEL === 'debug' || process.env.LOG_LEVEL === 'trace';
const logLines = [
`=== Container Run Log ===`,
`Timestamp: ${new Date().toISOString()}`,
`Group: ${group.name}`,
`IsMain: ${input.isMain}`,
`Duration: ${duration}ms`,
`Exit Code: ${code}`,
`Stdout Truncated: ${stdoutTruncated}`,
`Stderr Truncated: ${stderrTruncated}`,
``,
];
const isError = code !== 0;
if (isVerbose || isError) {
logLines.push(
`=== Input ===`,
JSON.stringify(input, null, 2),
``,
`=== Container Args ===`,
containerArgs.join(' '),
``,
`=== Mounts ===`,
mounts
.map(
(m) =>
`${m.hostPath} -> ${m.containerPath}${m.readonly ? ' (ro)' : ''}`,
)
.join('\n'),
``,
`=== Stderr${stderrTruncated ? ' (TRUNCATED)' : ''} ===`,
stderr,
``,
`=== Stdout${stdoutTruncated ? ' (TRUNCATED)' : ''} ===`,
stdout,
);
} else {
logLines.push(
`=== Input Summary ===`,
`Prompt length: ${input.prompt.length} chars`,
`Session ID: ${input.sessionId || 'new'}`,
``,
`=== Mounts ===`,
mounts
.map((m) => `${m.containerPath}${m.readonly ? ' (ro)' : ''}`)
.join('\n'),
``,
);
}
fs.writeFileSync(logFile, logLines.join('\n'));
logger.debug({ logFile, verbose: isVerbose }, 'Container log written');
if (code !== 0) {
logger.error(
{
group: group.name,
code,
duration,
stderr,
stdout,
logFile,
},
'Container exited with error',
);
resolve({
status: 'error',
result: null,
error: `Container exited with code ${code}: ${stderr.slice(-200)}`,
});
return;
}
// Streaming mode: wait for output chain to settle, return completion marker
if (onOutput) {
outputChain.then(() => {
logger.info(
{ group: group.name, duration, newSessionId },
'Container completed (streaming mode)',
);
resolve({
status: 'success',
result: null,
newSessionId,
});
});
return;
}
// Legacy mode: parse the last output marker pair from accumulated stdout
try {
// Extract JSON between sentinel markers for robust parsing
const startIdx = stdout.indexOf(OUTPUT_START_MARKER);
const endIdx = stdout.indexOf(OUTPUT_END_MARKER);
let jsonLine: string;
if (startIdx !== -1 && endIdx !== -1 && endIdx > startIdx) {
jsonLine = stdout
.slice(startIdx + OUTPUT_START_MARKER.length, endIdx)
.trim();
} else {
// Fallback: last non-empty line (backwards compatibility)
const lines = stdout.trim().split('\n');
jsonLine = lines[lines.length - 1];
}
const output: ContainerOutput = JSON.parse(jsonLine);
logger.info(
{
group: group.name,
duration,
status: output.status,
hasResult: !!output.result,
},
'Container completed',
);
resolve(output);
} catch (err) {
logger.error(
{
group: group.name,
stdout,
stderr,
error: err,
},
'Failed to parse container output',
);
resolve({
status: 'error',
result: null,
error: `Failed to parse container output: ${err instanceof Error ? err.message : String(err)}`,
});
}
});
container.on('error', (err) => {
clearTimeout(timeout);
logger.error(
{ group: group.name, containerName, error: err },
'Container spawn error',
);
resolve({
status: 'error',
result: null,
error: `Container spawn error: ${err.message}`,
});
});
});
}
export function writeTasksSnapshot(
groupFolder: string,
isMain: boolean,
tasks: Array<{
id: string;
groupFolder: string;
prompt: string;
schedule_type: string;
schedule_value: string;
status: string;
next_run: string | null;
}>,
): void {
// Write filtered tasks to the group's IPC directory
const groupIpcDir = resolveGroupIpcPath(groupFolder);
fs.mkdirSync(groupIpcDir, { recursive: true });
// Main sees all tasks, others only see their own
const filteredTasks = isMain
? tasks
: tasks.filter((t) => t.groupFolder === groupFolder);
const tasksFile = path.join(groupIpcDir, 'current_tasks.json');
fs.writeFileSync(tasksFile, JSON.stringify(filteredTasks, null, 2));
}
export interface AvailableGroup {
jid: string;
name: string;
lastActivity: string;
isRegistered: boolean;
}
/**
* Write available groups snapshot for the container to read.
* Only main group can see all available groups (for activation).
* Non-main groups only see their own registration status.
*/
export function writeGroupsSnapshot(
groupFolder: string,
isMain: boolean,
groups: AvailableGroup[],
registeredJids: Set<string>,
): void {
const groupIpcDir = resolveGroupIpcPath(groupFolder);
fs.mkdirSync(groupIpcDir, { recursive: true });
// Main sees all groups; others see nothing (they can't activate groups)
const visibleGroups = isMain ? groups : [];
const groupsFile = path.join(groupIpcDir, 'available_groups.json');
fs.writeFileSync(
groupsFile,
JSON.stringify(
{
groups: visibleGroups,
lastSync: new Date().toISOString(),
},
null,
2,
),
);
}

View File

@@ -0,0 +1,33 @@
# Intent: src/container-runner.ts modifications
## What changed
Updated `buildContainerArgs` to support Apple Container's .env shadowing mechanism. The function now accepts an `isMain` parameter and uses it to decide how container user identity is configured.
## Why
Apple Container (VirtioFS) only supports directory mounts, not file mounts. The previous approach of mounting `/dev/null` over `.env` from the host causes a `VZErrorDomain` crash. Instead, main-group containers now start as root so the entrypoint can `mount --bind /dev/null` over `.env` inside the Linux VM, then drop to the host user via `setpriv`.
## Key sections
### buildContainerArgs (signature change)
- Added: `isMain: boolean` parameter
- Main containers: passes `RUN_UID`/`RUN_GID` env vars instead of `--user`, so the container starts as root
- Non-main containers: unchanged, still uses `--user` flag
### buildVolumeMounts
- Removed: the `/dev/null``/workspace/project/.env` shadow mount (was in the committed `37228a9` fix)
- The .env shadowing is now handled inside the container entrypoint instead
### runContainerAgent (call site)
- Changed: `buildContainerArgs(mounts, containerName)``buildContainerArgs(mounts, containerName, input.isMain)`
## Invariants
- All exported interfaces unchanged: `ContainerInput`, `ContainerOutput`, `runContainerAgent`, `writeTasksSnapshot`, `writeGroupsSnapshot`, `AvailableGroup`
- Non-main containers behave identically (still get `--user` flag)
- Mount list for non-main containers is unchanged
- Secrets still passed via stdin, never mounted as files
- Output parsing (streaming + legacy) unchanged
## Must-keep
- The `isMain` parameter on `buildContainerArgs` (consumed by `runContainerAgent`)
- The `RUN_UID`/`RUN_GID` env vars for main containers (consumed by entrypoint.sh)
- The `--user` flag for non-main containers (file permission compatibility)