Xmz Sandbox Runtime (srt)

A lightweight sandboxing tool for enforcing filesystem and network restrictions on arbitrary processes at the OS level, without requiring a container.

srt uses native OS sandboxing primitives (sandbox-exec on macOS, bubblewrap on Linux) and proxy-based network filtering. It can be used to sandbox the behaviour of agents, local MCP servers, bash commands and arbitrary processes.

Note

This project is forked from Anthropic's Sandbox Runtime with significant architectural changes and enhancements. The API has been refactored from a singleton pattern to a flexible class-based architecture that supports multiple concurrent sandbox instances with different configurations.

Key Differences from Original

This fork introduces several major enhancements based on commit 9c0d868:

Architecture Changes

Class-based architecture: Refactored from singleton pattern to instantiable classes (NetworkManager and SandboxManager)
Dual-mode API: Simple mode (SandboxManager manages NetworkManager internally) and Advanced mode (share NetworkManager across workers)
Multi-worker support: Multiple concurrent sandbox instances with different filesystem/environment configurations
Separated concerns: Network configuration (proxy lifecycle) now separate from instance configuration (filesystem/env)

New Features

Custom environment variables (e647549): Set or inherit environment variables in sandboxed processes
Node.js global proxy (ccbe204): Automatic proxy configuration for Node.js HTTP clients via global-agent
WebSocket support (2cc0632): Handle WebSocket upgrade requests in HTTP proxy
Linux glob pattern support: Linux now supports glob patterns in filesystem paths (expanded via ripgrep at initialization)

Bug Fixes

Empty allowedDomains fix: Fixed security vulnerability where allowedDomains: [] didn't block network access
Non-existent path handling (71d9033): Improved write-deny for non-existent paths
Device file normalization: Skip normalization for /dev/* paths to preserve device access

Installation

npm install @xmz-ai/sandbox-runtime

Or install globally for CLI usage:

npm install -g @xmz-ai/sandbox-runtime

Basic Usage

# Network restrictions
$ srt "curl anthropic.com"
Running: curl anthropic.com
<html>...</html>  # Request succeeds

$ srt "curl example.com"
Running: curl example.com
Connection blocked by network allowlist  # Request blocked

# Filesystem restrictions
$ srt "cat README.md"
Running: cat README.md
# Anthropic Sandb...  # Current directory access allowed

$ srt "cat ~/.ssh/id_rsa"
Running: cat ~/.ssh/id_rsa
cat: /Users/ollie/.ssh/id_rsa: Operation not permitted  # Specific file blocked

Overview

This package provides a standalone sandbox implementation that can be used as both a CLI tool and a library. It's designed with a secure-by-default philosophy tailored for common developer use cases: processes start with minimal access, and you explicitly poke only the holes you need.

Key capabilities:

Network restrictions: Control which hosts/domains can be accessed via HTTP/HTTPS and other protocols
Filesystem restrictions: Control which files/directories can be read/written
Unix socket restrictions: Control access to local IPC sockets
Violation monitoring: On macOS, tap into the system's sandbox violation log store for real-time alerts

Example Use Case: Sandboxing MCP Servers

A key use case is sandboxing Model Context Protocol (MCP) servers to restrict their capabilities. For example, to sandbox the filesystem MCP server:

Without sandboxing (.mcp.json):

{
  "mcpServers": {
    "filesystem": {
      "command": "npx",
      "args": ["-y", "@modelcontextprotocol/server-filesystem"]
    }
  }
}

With sandboxing (.mcp.json):

{
  "mcpServers": {
    "filesystem": {
      "command": "srt",
      "args": ["npx", "-y", "@modelcontextprotocol/server-filesystem"]
    }
  }
}

Then configure restrictions in ~/.srt-settings.json:

{
  "filesystem": {
    "denyRead": [],
    "allowWrite": ["."],
    "denyWrite": ["~/sensitive-folder"]
  },
  "network": {
    "allowedDomains": [],
    "deniedDomains": []
  }
}

Now the MCP server will be blocked from writing to the denied path:

> Write a file to ~/sensitive-folder
✗ Error: EPERM: operation not permitted, open '/Users/ollie/sensitive-folder/test.txt'

How It Works

The sandbox uses OS-level primitives to enforce restrictions that apply to the entire process tree:

macOS: Uses sandbox-exec with dynamically generated Seatbelt profiles
Linux: Uses bubblewrap for containerization with network namespace isolation

Dual Isolation Model

Both filesystem and network isolation are required for effective sandboxing. Without file isolation, a compromised process could exfiltrate SSH keys or other sensitive files. Without network isolation, a process could escape the sandbox and gain unrestricted network access.

Filesystem Isolation enforces read and write restrictions:

Read: Supports two modes depending on platform and configuration:
- Deny-only mode (default on macOS, optional on Linux): By default, read access is allowed everywhere. You can deny specific paths using denyRead (e.g., ~/.ssh). An empty deny list means full read access.
- Allow-only mode (Linux only): By default, read access is denied everywhere. You must explicitly allow paths using allowRead (e.g., /usr, .). System paths like /usr, /bin are auto-included unless disabled. An empty allow list means no read access (except system paths if enabled). Use denyRead to block specific paths within allowed paths.
Write (allow-only pattern): By default, write access is denied everywhere. You must explicitly allow paths (e.g., ., /tmp). An empty allow list means no write access.

Network Isolation (allow-only pattern): By default, all network access is denied. You must explicitly allow domains. An empty allowedDomains list means no network access. Network traffic is routed through proxy servers running on the host:

Linux: Requests are routed via the filesystem over a Unix domain socket. The network namespace of the sandboxed process is removed entirely, so all network traffic must go through the proxies running on the host (listening on Unix sockets that are bind-mounted into the sandbox)
macOS: The Seatbelt profile allows communication only to a specific localhost port. The proxies listen on this port, creating a controlled channel for all network access

Both HTTP/HTTPS (via HTTP proxy) and other TCP traffic (via SOCKS5 proxy) are mediated by these proxies, which enforce your domain allowlists and denylists.

Changes from Original Architecture:

This fork has refactored the singleton-based architecture into a class-based system:

Before: SandboxManager.initialize() / SandboxManager.reset() (static methods, global state)
After: new SandboxManager() / sandbox.initialize() (instance methods, no global state)

This enables multiple concurrent sandbox instances with different configurations, which was not possible with the original singleton design.

For more details on the original sandboxing approach in Claude Code, see:

Architecture

src/
├── index.ts                  # Library exports
├── cli.ts                    # CLI entrypoint (srt command)
├── utils/                    # Shared utilities
│   ├── debug.ts             # Debug logging
│   ├── settings.ts          # Settings reader (permissions + sandbox config)
│   ├── platform.ts          # Platform detection
│   └── exec.ts              # Command execution utilities
└── sandbox/                  # Sandbox implementation
    ├── network-manager.ts    # NetworkManager class (proxy lifecycle)
    ├── sandbox-manager.ts    # SandboxManager class (per-instance restrictions)
    ├── sandbox-config.ts     # Zod schemas and TypeScript types
    ├── sandbox-schemas.ts    # Internal restriction schemas
    ├── sandbox-violation-store.ts # Violation tracking
    ├── sandbox-utils.ts      # Shared sandbox utilities
    ├── sandbox-dependencies.ts # Dependency checking
    ├── http-proxy.ts         # HTTP/HTTPS proxy for network filtering
    ├── socks-proxy.ts        # SOCKS5 proxy for network filtering
    ├── linux-sandbox-utils.ts # Linux bubblewrap sandboxing
    ├── linux-network-bridge.ts # Linux Unix socket bridge (socat)
    └── macos-sandbox-utils.ts # macOS sandbox-exec sandboxing (Seatbelt profiles)

Key Components:

NetworkManager: Manages HTTP/SOCKS proxy servers for network filtering. Can be shared across multiple SandboxManager instances or created per-instance.
SandboxManager: Manages filesystem/environment restrictions for a specific sandbox instance. Multiple instances can run concurrently with different configurations.
Platform-specific utilities: Linux uses bubblewrap for containerization, macOS uses sandbox-exec with dynamically generated Seatbelt profiles.

Usage

As a CLI tool

The srt command (Anthropic Sandbox Runtime) wraps any command with security boundaries:

# Run a command in the sandbox
srt echo "hello world"

# With debug logging
srt --debug curl https://example.com

# Specify custom settings file
srt --settings /path/to/srt-settings.json npm install

As a library

Simple Mode: Single Sandbox Instance

For single-process applications or when you don't need to share network proxies:

import { SandboxManager } from '@xmz-ai/sandbox-runtime'
import { spawn } from 'child_process'

// Create SandboxManager with network and filesystem configuration
// SandboxManager will create and manage its own NetworkManager internally
const sandbox = new SandboxManager(
  // Network config
  {
    allowedDomains: ['example.com', 'api.github.com'],
    deniedDomains: [],
  },
  // Sandbox instance config
  {
    filesystem: {
      denyRead: ['~/.ssh'],
      allowWrite: ['.', '/tmp'],
      denyWrite: ['.env'],
    },
  },
)

// Initialize network proxies (required in simple mode)
await sandbox.initialize()

// Wrap a command with sandbox restrictions
const sandboxedCommand = await sandbox.wrapWithSandbox(
  'curl https://example.com',
)

// Execute the sandboxed command
const child = spawn(sandboxedCommand, { shell: true, stdio: 'inherit' })

// Handle exit
child.on('exit', async code => {
  console.log(`Command exited with code ${code}`)

  // Cleanup - automatically shuts down network proxies
  await sandbox.dispose()
})

Advanced Mode: Multiple Workers with Shared Network Proxy

For daemon processes that spawn multiple workers with different configurations but shared network filtering:

import { NetworkManager, SandboxManager } from '@xmz-ai/sandbox-runtime'

// Create shared NetworkManager (used by all workers)
const networkManager = new NetworkManager()
await networkManager.initialize({
  allowedDomains: ['github.com', '*.npmjs.org'],
  deniedDomains: [],
})

// Create Worker 1 with specific filesystem config
const worker1Sandbox = new SandboxManager(networkManager, {
  filesystem: {
    allowWrite: ['/workspace/project1'],
  },
  env: {
    NODE_ENV: 'development',
  },
})

// Create Worker 2 with different filesystem config
const worker2Sandbox = new SandboxManager(networkManager, {
  filesystem: {
    allowWrite: ['/workspace/project2'],
  },
  env: {
    NODE_ENV: 'production',
  },
})

// Each worker can execute commands with its own restrictions
const cmd1 = await worker1Sandbox.wrapWithSandbox('npm install')
const cmd2 = await worker2Sandbox.wrapWithSandbox('python build.py')

// Cleanup - dispose workers first, then shutdown shared network
worker1Sandbox.dispose() // Does NOT shutdown network (not owned)
worker2Sandbox.dispose() // Does NOT shutdown network (not owned)
await networkManager.shutdown() // Manually shutdown shared network proxies

Key differences between modes:

Simple mode: Pass NetworkConfig as first parameter. SandboxManager creates and manages its own NetworkManager. Call sandbox.initialize() before use, and sandbox.dispose() shuts down both sandbox and network.
Advanced mode: Pass NetworkManager instance as first parameter. You control when to create and shutdown the NetworkManager. Multiple SandboxManager instances can share one NetworkManager. sandbox.dispose() only cleans up the instance, not the shared NetworkManager.

Available Exports

// Core classes
import {
  NetworkManager, // Manages HTTP/SOCKS proxies for network filtering
  SandboxManager, // Manages sandbox instances with filesystem/env restrictions
  SandboxViolationStore, // Tracks sandbox violations
} from '@xmz-ai/sandbox-runtime'

// Configuration types
import type {
  NetworkConfig, // Network configuration (allowedDomains, etc.)
  SandboxInstanceConfig, // Per-instance config (filesystem, env)
  SandboxOptions, // Optional constructor options
  FilesystemConfig, // Filesystem restrictions
  IgnoreViolationsConfig, // Violation filtering rules
  EnvConfig, // Environment variable config
  RipgrepConfig, // Custom ripgrep configuration
} from '@xmz-ai/sandbox-runtime'

// Internal types (for advanced usage)
import type {
  NetworkContext, // Network context (proxy ports, Linux bridges)
  FsReadRestrictionConfig, // Internal read restriction format
  FsWriteRestrictionConfig, // Internal write restriction format
} from '@xmz-ai/sandbox-runtime'

Constructor Signatures:

// Simple mode
new SandboxManager(
  networkConfig: NetworkConfig,
  instanceConfig: SandboxInstanceConfig,
  options?: SandboxOptions
)

// Advanced mode
new SandboxManager(
  networkManager: NetworkManager,
  instanceConfig: SandboxInstanceConfig,
  options?: SandboxOptions
)

Configuration

Settings File Location

By default, the sandbox runtime looks for configuration at ~/.srt-settings.json. You can specify a custom path using the --settings flag:

srt --settings /path/to/srt-settings.json <command>

Configuration Reference

NetworkConfig (for NetworkManager or simple mode)

interface NetworkConfig {
  allowedDomains: string[] // Allowed domains (supports wildcards like "*.example.com")
  deniedDomains: string[] // Denied domains (checked first, takes precedence)
  allowUnixSockets?: string[] // Unix socket paths to allow (macOS only)
  allowAllUnixSockets?: boolean // Allow all Unix sockets (Linux only)
  allowLocalBinding?: boolean // Allow binding to local ports (default: false)
  httpProxyPort?: number // Use external HTTP proxy instead of starting one
  socksProxyPort?: number // Use external SOCKS proxy instead of starting one
}

SandboxInstanceConfig (for SandboxManager)

interface SandboxInstanceConfig {
  filesystem: FilesystemConfig
  env?: Record<string, string | null> // Environment variables (null = inherit from host)
  ignoreViolations?: Record<string, string[]> // Violation filtering rules
  enableWeakerNestedSandbox?: boolean // Weaker mode for Docker (Linux only)
  ripgrep?: { command: string; args?: string[] } // Custom ripgrep config
  mandatoryDenySearchDepth?: number // Max depth for dangerous file search (1-10, default: 3)
  allowPty?: boolean // Allow pseudo-terminal operations (macOS only, for tmux/screen)
}

SandboxOptions (optional constructor parameter)

interface SandboxOptions {
  enableLogMonitor?: boolean // Enable macOS sandbox log monitoring (default: false)
}

Complete Configuration Example

Example 1: Standard mode with simplified domain patterns:

{
  "network": {
    "allowedDomains": [
      ".github.com",  // Matches github.com AND all subdomains (simplified!)
      ".npmjs.org"    // Matches npmjs.org AND all subdomains
    ],
    "deniedDomains": ["malicious.com"],
    "allowUnixSockets": ["/var/run/docker.sock"],
    "allowLocalBinding": false
  },
  "filesystem": {
    "denyRead": ["~/.ssh"],
    "allowWrite": [".", "src/", "test/", "/tmp"],
    "denyWrite": [".env", "config/production.json"]
  },
  "ignoreViolations": {
    "*": ["/usr/bin", "/System"],
    "git push": ["/usr/bin/nc"],
    "npm": ["/private/tmp"]
  },
  "allowPty": false,
  "enableWeakerNestedSandbox": false
}

Example 2: Allow-all mode for development:

{
  "network": {
    "allowedDomains": "*",  // Allow all domains (string, not array)
    "deniedDomains": [".internal-corp.com", "malicious.com"]  // Block internal and known bad domains
  },
  "filesystem": {
    "denyRead": ["~/.ssh"],
    "allowWrite": ["."],
    "denyWrite": [".env"]
  }
}

Example 3: Deny-all mode with strict whitelist (Linux only):

{
  "network": {
    "allowedDomains": ["github.com", "npmjs.org"],  // Only these domains allowed
    "deniedDomains": "*"  // Deny all by default (string, not array)
  },
  "filesystem": {
    "allowRead": [".", "src/", "test/", "/tmp"],
    "denyRead": [".env", "secrets.json", ".aws/credentials"],
    "autoAllowSystemPaths": true,
    "allowWrite": [".", "src/", "test/", "/tmp"],
    "denyWrite": [".env", ".git"]
  },
  "allowPty": false,
  "mandatoryDenySearchDepth": 3,
  "enableWeakerNestedSandbox": false
}

Example with PTY support for tmux/screen (macOS only):

{
  "network": {
    "allowedDomains": ["github.com"],
    "deniedDomains": []
  },
  "filesystem": {
    "allowWrite": ["."]
  },
  "allowPty": true
}

When allowPty is true on macOS, the sandbox allows pseudo-terminal operations needed by terminal multiplexers like tmux, screen, and the script command. This is disabled by default for security.

Configuration Options

Network Configuration

Uses flexible filtering modes with multiple pattern types.

Configuration Fields:

network.allowedDomains - Allowed domain patterns. Can be:
- "*" (string) - Allow-all mode: permits all domains except those in deniedDomains
- Array of patterns - Standard mode: only listed patterns are allowed
network.deniedDomains - Denied domain patterns. Can be:
- "*" (string) - Deny-all mode: blocks all domains except those in allowedDomains
- Array of patterns - Block specific domains (takes precedence over allowedDomains)
network.allowUnixSockets - Array of Unix socket paths that can be accessed (macOS only)
network.allowLocalBinding - Allow binding to local ports (boolean, default: false)

Domain Pattern Types:

Exact match: "example.com" - matches only example.com
Subdomain wildcard: "*.example.com" - matches api.example.com, sub.api.example.com, but NOT example.com
Full wildcard: ".example.com" - matches example.com AND all subdomains (e.g., api.example.com, sub.api.example.com)
Match all: "*" - matches any domain (must be a string, not in array)
localhost: "localhost" - matches localhost

Filtering Modes:

Standard mode (default): allowedDomains is an array
- Default: deny all
- Check deniedDomains first (if matches, deny)
- Then check allowedDomains (if matches, allow)
- Otherwise, deny
Allow-all mode: allowedDomains: "*"
- Default: allow all
- Only deniedDomains can block access
- Useful for development or when network filtering is handled externally
Deny-all mode: deniedDomains: "*"
- Default: deny all
- Only allowedDomains can permit access
- Strictest mode: explicit whitelist only

Filesystem Configuration

Read restrictions support two modes:

Deny-only mode (backward compatible):

filesystem.denyRead - Array of paths to deny read access. By default, all paths are readable except the denied ones.
Supported on: Linux and macOS
Use case: When you want to allow broad read access but block specific sensitive paths (e.g., ~/.ssh, credentials files)
Example: "denyRead": ["~/.ssh", "~/.aws"] means everything is readable except those paths

Allow-only mode (more secure):

filesystem.allowRead - Array of paths to allow read access. By default, only specified paths are readable.
filesystem.denyRead - When used with allowRead, this denies specific paths within the allowed paths (deny-within-allow pattern).
filesystem.autoAllowSystemPaths - Boolean (default: true). Automatically include system paths (like /usr, /bin, /lib) needed for commands to execute.
Supported on: Linux only (macOS does not support this mode due to sandbox-exec limitations)
Use case: When you want maximum security and explicit control over what can be read
Example: "allowRead": [".", "/tmp"], "denyRead": [".env", "secrets.json"] with autoAllowSystemPaths: true means only those paths (plus system paths) are readable, except .env and secrets.json are blocked

Important notes:

denyRead semantics: The meaning of denyRead depends on whether you use allowRead:
- Without allowRead (deny-only mode): denyRead globally denies paths
- With allowRead (allow-only mode): denyRead denies paths within allowed paths
macOS limitation: macOS only supports deny-only mode. If you specify allowRead on macOS, an error will be thrown
Linux supports both: On Linux, you can use denyRead alone (deny-only) or combine allowRead + denyRead (allow-only with exceptions)
Default behavior: If neither is specified, no read restrictions are applied (all reads allowed)

Write restrictions (allow-only pattern) - all writes denied by default:

filesystem.allowWrite - Array of paths to allow write access. Empty array = no write access.
filesystem.denyWrite - Array of paths to deny write access within allowed paths (takes precedence over allowWrite)

Path Syntax (Both Platforms):

Both macOS and Linux support git-style glob patterns, similar to .gitignore syntax:

* - Matches any characters except / (e.g., *.ts matches foo.ts but not foo/bar.ts)
** - Matches any characters including / (e.g., src/**/*.ts matches all .ts files in src/)
? - Matches any single character except / (e.g., file?.txt matches file1.txt)
[abc] - Matches any character in the set (e.g., file[0-9].txt matches file3.txt)

Examples:

"allowWrite": ["src/"] - Allow write to entire src/ directory
"allowWrite": ["src/**/*.ts"] - Allow write to all .ts files in src/ and subdirectories
"denyRead": ["~/.ssh"] - Deny read to SSH directory
"denyWrite": ["**/.env"] - Deny write to all .env files recursively

Platform Differences:

macOS: Uses regex matching in sandbox profiles. Glob patterns protect both existing files AND files created in the future during sandbox execution.
Linux: Expands glob patterns at initialization time using ripgrep. Glob patterns only protect files that exist when the sandbox starts. Files created after initialization are NOT protected by the pattern.
- Use mandatoryDenySearchDepth (default: 3) to control how deep ripgrep searches for matching files
- Deeper searches provide better protection but may impact initialization performance

All platforms:

Paths can be absolute (e.g., /home/user/.ssh) or relative to the current working directory (e.g., ./src)
~ expands to the user's home directory

Other Configuration

ignoreViolations - Object mapping command patterns to arrays of paths where violations should be ignored (macOS only)
allowPty - Allow pseudo-terminal (PTY) operations for tmux and other terminal multiplexers (macOS only, default: false)
enableWeakerNestedSandbox - Enable weaker sandbox mode for Docker environments (Linux only, default: false)
mandatoryDenySearchDepth - Maximum directory depth to search for dangerous files (Linux only, 1-10, default: 3)
ripgrep - Custom ripgrep configuration for file scanning (Linux only)
env - Custom environment variables to set in sandboxed processes (string value or null to inherit from host)

Pattern Migration Guide

The new .example.com pattern simplifies configuration by matching both the base domain and all subdomains with a single rule.

Before (multiple rules needed):

{
  "allowedDomains": ["github.com", "*.github.com"]  // Two rules for base + subdomains
}

After (single rule):

{
  "allowedDomains": [".github.com"]  // One rule matches both!
}

Pattern comparison:

Pattern	Matches `example.com`	Matches `api.example.com`	Matches `deep.api.example.com`
`"example.com"`	✅ Yes	❌ No	❌ No
`"*.example.com"`	❌ No	✅ Yes	✅ Yes
`".example.com"` (NEW)	✅ Yes	✅ Yes	✅ Yes
`"*"` (NEW)	✅ Yes	✅ Yes	✅ Yes

Common Configuration Recipes

Allow GitHub access (simplified with new pattern):

{
  "network": {
    "allowedDomains": [".github.com"],  // Matches github.com AND all subdomains!
    "deniedDomains": []
  },
  "filesystem": {
    "denyRead": [],
    "allowWrite": ["."],
    "denyWrite": []
  }
}

Restrict to specific directories (deny-only mode):

{
  "network": {
    "allowedDomains": [],
    "deniedDomains": []
  },
  "filesystem": {
    "denyRead": ["~/.ssh"],
    "allowWrite": [".", "src/", "test/"],
    "denyWrite": [".env", "secrets/"]
  }
}

Maximum security: only allow reading specific directories (allow-only mode, Linux only):

{
  "network": {
    "allowedDomains": [],
    "deniedDomains": []
  },
  "filesystem": {
    "allowRead": [".", "/tmp"],
    "denyRead": [".env", "secrets.json"],
    "autoAllowSystemPaths": true,
    "allowWrite": ["."],
    "denyWrite": [".env", ".git"]
  }
}

This configuration (Linux only):

Only allows reading the current directory and /tmp
Blocks reading .env and secrets.json even within the current directory
Automatically includes system paths (/usr, /bin, /lib, etc.) for commands to execute
Only allows writing to the current directory
Blocks writing to .env and .git even within the current directory
Note: On macOS, this will fall back to deny-only mode (no read restrictions)

Sandbox MCP servers or AI agents (development mode):

{
  "network": {
    "allowedDomains": "*",  // Allow all for development
    "deniedDomains": [".internal-company.com"]  // Block internal domains
  },
  "filesystem": {
    "allowRead": ["/path/to/project"],
    "denyRead": ["/path/to/project/.env", "/path/to/project/secrets.json"],
    "autoAllowSystemPaths": true,
    "allowWrite": ["/path/to/project"],
    "denyWrite": ["/path/to/project/.env", "/path/to/project/.git"]
  },
  "env": {
    "NODE_ENV": "production",
    "PATH": null,
    "HOME": null
  }
}

This configuration:

Restricts access to a specific project directory
Blocks sensitive files like SSH keys, AWS credentials, and secrets
Sets NODE_ENV to a specific value
Inherits PATH and HOME from the host environment (null = inherit)

Common Issues and Tips

Running Jest: Use --no-watchman flag to avoid sandbox violations:

srt "jest --no-watchman"

Watchman accesses files outside the sandbox boundaries, which will trigger permission errors. Disabling it allows Jest to run with the built-in file watcher instead.

Platform Support

macOS: Uses sandbox-exec with custom profiles (no additional dependencies)
Linux: Uses bubblewrap (bwrap) for containerization
Windows: Not yet supported

Platform-Specific Dependencies

Linux requires:

bubblewrap - Container runtime
- Ubuntu/Debian: apt-get install bubblewrap
- Fedora: dnf install bubblewrap
- Arch: pacman -S bubblewrap
socat - Socket relay for proxy bridging
- Ubuntu/Debian: apt-get install socat
- Fedora: dnf install socat
- Arch: pacman -S socat
ripgrep - Fast search tool for deny path detection
- Ubuntu/Debian: apt-get install ripgrep
- Fedora: dnf install ripgrep
- Arch: pacman -S ripgrep

Optional Linux dependencies (for seccomp fallback):

The package includes pre-generated seccomp BPF filters for x86-64 and arm architectures. These dependencies are only needed if you are on a different architecture where pre-generated filters are not available:

gcc or clang - C compiler
libseccomp-dev - Seccomp library development files
- Ubuntu/Debian: apt-get install gcc libseccomp-dev
- Fedora: dnf install gcc libseccomp-devel
- Arch: pacman -S gcc libseccomp

macOS requires:

ripgrep - Fast search tool for deny path detection
- Install via Homebrew: brew install ripgrep
- Or download from: https://github.com/BurntSushi/ripgrep/releases

Note: On macOS, ripgrep is only required if you plan to use custom ripgrep configurations. The built-in rg command is sufficient for normal usage.

Development

# Install dependencies
npm install

# Build the project
npm run build

# Build seccomp binaries (requires Docker)
npm run build:seccomp

# Run tests
npm test

# Run integration tests
npm run test:integration

# Type checking
npm run typecheck

# Lint code
npm run lint

# Format code
npm run format

Building Seccomp Binaries

The pre-generated BPF filters are included in the repository, but you can rebuild them if needed:

npm run build:seccomp

This script uses Docker to cross-compile seccomp binaries for multiple architectures:

x64 (x86-64)
arm64 (aarch64)

The script builds static generator binaries, generates the BPF filters (~104 bytes each), and stores them in vendor/seccomp/x64/ and vendor/seccomp/arm64/. The generator binaries are removed to keep the package size small.

Implementation Details

Network Isolation Architecture

The sandbox runs HTTP and SOCKS5 proxy servers on the host machine that filter all network requests based on permission rules:

HTTP/HTTPS Traffic: An HTTP proxy server intercepts requests and validates them against allowed/denied domains
Other Network Traffic: A SOCKS5 proxy handles all other TCP connections (SSH, database connections, etc.)
Permission Enforcement: The proxies enforce the permissions rules from your configuration

Platform-specific proxy communication:

Linux: Requests are routed via the filesystem over Unix domain sockets (using socat for bridging). The network namespace is removed from the bubblewrap container, ensuring all network traffic must go through the proxies.
macOS: The Seatbelt profile allows communication only to specific localhost ports where the proxies listen. All other network access is blocked.

Filesystem Isolation

Filesystem restrictions are enforced at the OS level:

macOS: Uses sandbox-exec with dynamically generated Seatbelt profiles that specify allowed read/write paths
Linux: Uses bubblewrap with bind mounts, marking directories as read-only or read-write based on configuration

Default filesystem permissions:

Read (deny-only): Allowed everywhere by default. You can deny specific paths.
- Example: denyRead: ["~/.ssh"] to block access to SSH keys
- Empty denyRead: [] = full read access (nothing denied)
Write (allow-only): Denied everywhere by default. You must explicitly allow paths.
- Example: allowWrite: [".", "/tmp"] to allow writes to current directory and /tmp
- Empty allowWrite: [] = no write access (nothing allowed)
- denyWrite creates exceptions within allowed paths

This model lets you start with broad read access but maximally restricted write access, then explicitly open the holes you need.

Protecting Sensitive Files

The sandbox does NOT automatically protect any files - you have full control via the denyWrite configuration. However, you should consider protecting sensitive files that could lead to sandbox escapes or system compromise:

Recommended files to protect:

{
  "filesystem": {
    "allowWrite": ["."],
    "denyWrite": [
      "**/.bashrc",
      "**/.bash_profile",
      "**/.zshrc",
      "**/.zprofile",
      "**/.profile",
      "**/.gitconfig",
      "**/.gitmodules",
      "**/.ripgreprc",
      "**/.mcp.json",
      "**/.git/hooks/**",
      "**/.git/config",
      "**/.vscode/**",
      "**/.idea/**",
      "**/.claude/commands/**",
      "**/.claude/agents/**"
    ]
  }
}

Using glob patterns (like **/.bashrc) protects these files anywhere in your project:

$ srt 'echo "malicious" >> .bashrc'
/bin/bash: .bashrc: Operation not permitted

$ srt 'echo "bad" > src/.git/hooks/pre-commit'
/bin/bash: src/.git/hooks/pre-commit: Operation not permitted

Search depth for glob patterns (Linux only): On Linux, glob patterns are expanded at initialization using ripgrep. By default, it searches up to 3 directory levels deep for performance. You can configure this with mandatoryDenySearchDepth:

{
  "mandatoryDenySearchDepth": 5,
  "filesystem": {
    "allowWrite": ["."],
    "denyWrite": ["**/.env*"]
  }
}

Default: 3 (searches 3 levels deep: ./, ./foo/, ./foo/bar/)
Range: 1 to 10
Higher values provide more protection but may slow initialization
This setting only affects Linux glob expansion; macOS uses regex matching with no depth limit

Unix Socket Restrictions (Linux)

On Linux, the sandbox uses seccomp BPF (Berkeley Packet Filter) to block Unix domain socket creation at the syscall level. This provides an additional layer of security to prevent processes from creating new Unix domain sockets for local IPC (unless explicitly allowed).

How it works:

Pre-generated BPF filters: The package includes pre-compiled BPF filters for different architectures (x64, ARM64). These are ~104 bytes each and stored in vendor/seccomp/. The filters are architecture-specific but libc-independent, so they work with both glibc and musl.
Runtime detection: The sandbox automatically detects your system's architecture and loads the appropriate pre-generated BPF filter.
Syscall filtering: The BPF filter intercepts the socket() syscall and blocks creation of AF_UNIX sockets by returning EPERM. This prevents sandboxed code from creating new Unix domain sockets.
Two-stage application using apply-seccomp binary:
- Outer bwrap creates the sandbox with filesystem, network, and PID namespace restrictions
- Network bridging processes (socat) start inside the sandbox (need Unix sockets)
- apply-seccomp binary applies the seccomp filter via prctl()
- apply-seccomp execs the user command with seccomp active
- User command runs with all sandbox restrictions plus Unix socket creation blocking

Security limitations: The filter only blocks socket(AF_UNIX, ...) syscalls. It does not prevent operations on Unix socket file descriptors inherited from parent processes or passed via SCM_RIGHTS. For most sandboxing scenarios, blocking socket creation is sufficient to prevent unauthorized IPC.

Zero runtime dependencies: Pre-built static apply-seccomp binaries and pre-generated BPF filters are included for x64 and arm64 architectures. No compilation tools or external dependencies required at runtime.

Architecture support: x64 and arm64 are fully supported with pre-built binaries. Other architectures are not currently supported. To use sandboxing without Unix socket blocking on unsupported architectures, set allowAllUnixSockets: true in your configuration.

Violation Detection and Monitoring

When a sandboxed process attempts to access a restricted resource:

Blocks the operation at the OS level (returns EPERM error)
Logs the violation (platform-specific mechanisms)
Notifies the user (in Claude Code, this triggers a permission prompt)

macOS: The sandbox runtime taps into macOS's system sandbox violation log store. This provides real-time notifications with detailed information about what was attempted and why it was blocked. This is the same mechanism Claude Code uses for violation detection.

# View sandbox violations in real-time
log stream --predicate 'process == "sandbox-exec"' --style syslog

Linux: Bubblewrap doesn't provide built-in violation reporting. Use strace to trace system calls and identify blocked operations:

# Trace all denied operations
strace -f srt <your-command> 2>&1 | grep EPERM

# Trace specific file operations
strace -f -e trace=open,openat,stat,access srt <your-command> 2>&1 | grep EPERM

# Trace network operations
strace -f -e trace=network srt <your-command> 2>&1 | grep EPERM

Advanced: Bring Your Own Proxy

For more sophisticated network filtering, you can configure the sandbox to use your own proxy instead of the built-in ones. This enables:

Traffic inspection: Use tools like mitmproxy to inspect and modify traffic
Custom filtering logic: Implement complex rules beyond simple domain allowlists
Audit logging: Log all network requests for compliance or debugging

Example with mitmproxy:

# Start mitmproxy with custom filtering script
mitmproxy -s custom_filter.py --listen-port 8888

Note: Custom proxy configuration is not yet supported in the new configuration format. This feature will be added in a future release.

Important security consideration: Even with domain allowlists, exfiltration vectors may exist. For example, allowing github.com lets a process push to any repository. With a custom MITM proxy and proper certificate setup, you can inspect and filter specific API calls to prevent this.

Security Limitations

Network Sandboxing Limitations: The network filtering system operates by restricting the domains that processes are allowed to connect to. It does not otherwise inspect the traffic passing through the proxy and users are responsible for ensuring they only allow trusted domains in their policy.

Users should be aware of potential risks that come from allowing broad domains like `github.com` that may allow for data exfiltration. Also, in some cases it may be possible to bypass the network filtering through [domain fronting](https://en.wikipedia.org/wiki/Domain_fronting).

Privilege Escalation via Unix Sockets: The allowUnixSockets configuration can inadvertently grant access to powerful system services that could lead to sandbox bypasses. For example, if it is used to allow access to /var/run/docker.sock this would effectively grant access to the host system through exploiting the docker socket. Users are encouraged to carefully consider any unix sockets that they allow through the sandbox.
Filesystem Permission Escalation: Overly broad filesystem write permissions can enable privilege escalation attacks. Allowing writes to directories containing executables in $PATH, system configuration directories, or user shell configuration files (.bashrc, .zshrc) can lead to code execution in different security contexts when other users or system processes access these files.
Linux Sandbox Strength: The Linux implementation provides strong filesystem and network isolation but includes an enableWeakerNestedSandbox mode that enables it to work inside of Docker environments without privileged namespaces. This option considerably weakens security and should only be used incases where additional isolation is otherwise enforced.

Known Limitations and Future Work

Linux proxy bypass: Currently uses environment variables (HTTP_PROXY, HTTPS_PROXY, ALL_PROXY) to direct traffic through proxies. This works for most applications but may be ignored by programs that don't respect these variables, leading to them being unable to connect to the internet.

Future improvements:

Proxychains support: Add support for proxychains with LD_PRELOAD on Linux to intercept network calls at a lower level, making bypass more difficult
Linux violation monitoring: Implement automatic strace-based violation detection for Linux, integrated with the violation store. Currently, Linux users must manually run strace to see violations, unlike macOS which has automatic violation monitoring via the system log store

Name		Name	Last commit message	Last commit date
Latest commit History 127 Commits
.github/workflows		.github/workflows
.husky		.husky
.vscode		.vscode
scripts		scripts
src		src
test		test
vendor		vendor
.dockerignore		.dockerignore
.gitignore		.gitignore
.npmrc		.npmrc
.prettierrc.json		.prettierrc.json
LICENSE		LICENSE
README.md		README.md
eslint.config.js		eslint.config.js
package-lock.json		package-lock.json
package.json		package.json
tsconfig.json		tsconfig.json
tsconfig.test.json		tsconfig.test.json

License

xmz-ai/sandbox-runtime

Folders and files

Latest commit

History

Repository files navigation