sandboxed.sh
starsforks
How to install sandboxed.sh?
th0rgal/sandboxed.sh
$git clone https://github.com/Th0rgal/sandboxed.sh<br/>
README.md
| 1 | <p align="center"> |
| 2 | <img src="dashboard/public/favicon.svg" width="120" alt="sandboxed.sh" /> |
| 3 | </p> |
| 4 | |
| 5 | <h1 align="center">sandboxed.sh</h1> |
| 6 | |
| 7 | <p align="center"> |
| 8 | <strong>Self-hosted cloud orchestrator for AI coding agents</strong><br/> |
| 9 | Isolated Linux workspaces with Claude Code, OpenCode, Codex, Gemini, and Grok runtimes |
| 10 | </p> |
| 11 | |
| 12 | <p align="center"> |
| 13 | <em>Formerly known as Open Agent</em> |
| 14 | </p> |
| 15 | |
| 16 | <p align="center"> |
| 17 | <a href="https://sandboxed.sh">Website</a> · |
| 18 | <a href="https://relens.ai/community">Discord</a> · |
| 19 | <a href="#vision">Vision</a> · |
| 20 | <a href="#features">Features</a> · |
| 21 | <a href="#ecosystem">Ecosystem</a> · |
| 22 | <a href="#screenshots">Screenshots</a> · |
| 23 | <a href="#getting-started">Getting Started</a> |
| 24 | </p> |
| 25 | |
| 26 | <br/> |
| 27 | |
| 28 | <p align="center"> |
| 29 | <img src="screenshots/hero.webp" alt="sandboxed.sh Dashboard" width="100%" /> |
| 30 | </p> |
| 31 | |
| 32 | <p align="center"> |
| 33 | <strong>Ready to deploy?</strong> Jump to the <a href="#choose-your-installation-method">installation comparison</a>, or go straight to the <a href="docs/install-docker.md">Docker guide</a> / <a href="docs/install-native.md">native guide</a>. |
| 34 | </p> |
| 35 | |
| 36 | --- |
| 37 | |
| 38 | ## Vision |
| 39 | |
| 40 | What if you could: |
| 41 | |
| 42 | **Hand off entire dev cycles.** Point an agent at a GitHub issue, let it write |
| 43 | code, test by launching desktop applications, and open a PR when tests pass. You |
| 44 | review the diff, not the process. |
| 45 | |
| 46 | **Run multi-day operations unattended.** Give an agent SSH access to your home |
| 47 | GPU through a VPN. It reads Nvidia docs, sets up training, fine-tunes models |
| 48 | while you sleep. |
| 49 | |
| 50 | **Keep sensitive data local.** Analyze your sequenced DNA against scientific |
| 51 | literature. Local inference, isolated containers, nothing leaves your machines. |
| 52 | |
| 53 | --- |
| 54 | |
| 55 | ## Features |
| 56 | |
| 57 | - **Multi-Runtime Support**: Run Claude Code, OpenCode, Codex, Gemini, and Grok |
| 58 | agents in the same infrastructure |
| 59 | - **Mission Control**: Start, stop, and monitor agents remotely with real-time |
| 60 | streaming |
| 61 | - **Isolated Workspaces**: Containerized Linux environments (systemd-nspawn) |
| 62 | with per-mission directories |
| 63 | - **Git-backed Library**: Skills, tools, rules, agents, and MCPs versioned in a |
| 64 | single repo |
| 65 | - **Assistant Gateway**: Manage Telegram gateway compatibility from the |
| 66 | top-level Assistant UI while Hermes takes over assistant runtime over MCP |
| 67 | - **Automations**: Schedule recurring agent runs with cron-like triggers |
| 68 | - **Model Routing**: Provider fallback chains with health checks and |
| 69 | rate-limit handling |
| 70 | - **MCP Registry (optional)**: Extra tool servers (desktop/playwright/etc.) when |
| 71 | needed |
| 72 | - **OpenAI-compatible Proxy Queue Mode**: Optional deferred execution for |
| 73 | `/v1/chat/completions` when all routed providers are temporarily rate-limited |
| 74 | - **Multi-platform**: Web dashboard (Next.js) and iOS app (SwiftUI) with |
| 75 | Picture-in-Picture |
| 76 | |
| 77 | --- |
| 78 | |
| 79 | ## Ecosystem |
| 80 | |
| 81 | sandboxed.sh orchestrates multiple AI coding agent runtimes: |
| 82 | |
| 83 | - **[Claude Code](https://docs.anthropic.com/en/docs/claude-code)**: Anthropic's |
| 84 | official coding agent with native skills support (`.claude/skills/`) |
| 85 | - **[OpenCode](https://github.com/anomalyco/opencode)**: Open-source coding agent |
| 86 | - **Codex, Gemini, and Grok**: Native CLI backends for OpenAI, Google, and xAI |
| 87 | coding agents |
| 88 | |
| 89 | Each runtime executes inside isolated workspaces, so bash commands and file |
| 90 | operations are scoped correctly. sandboxed.sh handles orchestration, workspace |
| 91 | isolation, and Library-based configuration management. |
| 92 | |
| 93 | --- |
| 94 | |
| 95 | ## Screenshots |
| 96 | |
| 97 | <p align="center"> |
| 98 | <img src="screenshots/dashboard-overview.webp" alt="Dashboard Overview" width="100%" /> |
| 99 | </p> |
| 100 | <p align="center"><em>Real-time monitoring with CPU, memory, network graphs and mission timeline</em></p> |
| 101 | |
| 102 | <br/> |
| 103 | |
| 104 | <p align="center"> |
| 105 | <img src="screenshots/library-skills.webp" alt="Library Skills Editor" width="100%" /> |
| 106 | </p> |
| 107 | <p align="center"><em>Git-backed Library with skills, commands, rules, and inline editing</em></p> |
| 108 | |
| 109 | <br/> |
| 110 | |
| 111 | <p align="center"> |
| 112 | <img src="screenshots/mcp-servers.webp" alt="MCP Servers" width="100%" /> |
| 113 | </p> |
| 114 | <p align="center"><em>MCP server management with runtime status a |
People who liked this also viewed
6 picksType
- mcp-scannerScan MCP servers for potential threats & security findings.MCP ServersJul 2026
- hol-guardOpen-source antivirus for AI agents: block risky tools, secret access, prompt injection, malicious packages, MCP servers, plugins, and skills at runtime.MCP ServersJul 2026
- agent-scanSecurity scanner for AI agents, MCP servers and agent skills.MCP ServersJul 2026
- falcon-mcpConnect AI agents to CrowdStrike Falcon for automated security analysis and threat huntingMCP ServersJul 2026
- python🔐 Plug-and-play auth for Python MCP servers.MCP ServersAug 2025
- auth0-mcp-serverConnect LLMs and AI agents to Auth0 Management APIs via an MCP server. Use natural-language commands to create apps, deploy Actions, and query logs—runs locally and supports interactive tenant authentication for secure Auth0 administration from AI tools/CLIs.MCP ServersJul 2026
