by Wide-Moat
MCP server that gives any LLM its own computer — managed Docker workspaces with live browser, terminal, code execution, document skills, and autonomous sub-agents. Self-hosted, open-source, pluggable into any model.
# Add to your Claude Code skills
git clone https://github.com/Wide-Moat/open-computer-useGuides for using ai agents skills like open-computer-use.
Last scanned: 6/17/2026
{
"issues": [
{
"type": "npm-audit",
"message": "form-data: form-data uses unsafe random function in form-data for choosing boundary",
"severity": "critical"
},
{
"type": "npm-audit",
"message": "glob: glob CLI: Command injection via -c/--cmd executes matches with shell:true",
"severity": "high"
},
{
"type": "npm-audit",
"message": "gray-matter: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "js-yaml: js-yaml has prototype pollution in merge (<<)",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "markdown-it: markdown-it is has a Regular Expression Denial of Service (ReDoS)",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "markdown-pdf: markdown-pdf vulnerable to local file read via server side cross-site scripting (XSS)",
"severity": "high"
},
{
"type": "npm-audit",
"message": "markdown-toc: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "markdownlint-cli: Vulnerability found",
"severity": "high"
},
{
"type": "npm-audit",
"message": "markdownlint-cli2: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "minimatch: minimatch has a ReDoS via repeated wildcards with non-matching literal in pattern",
"severity": "high"
},
{
"type": "npm-audit",
"message": "phantomjs-prebuilt: Vulnerability found",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "qs: qs's arrayLimit bypass in its bracket notation allows DoS via memory exhaustion",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "request: Server-Side Request Forgery in Request",
"severity": "critical"
},
{
"type": "npm-audit",
"message": "smol-toml: smol-toml: Denial of Service via TOML documents containing thousands of consecutive commented lines",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "tmp: tmp allows arbitrary temporary file / directory write via symbolic link `dir` parameter",
"severity": "high"
},
{
"type": "npm-audit",
"message": "tough-cookie: tough-cookie Prototype Pollution vulnerability",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "uuid: uuid: Missing buffer bounds check in v3/v5/v6 when buf is provided",
"severity": "medium"
}
],
"status": "FAILED",
"scannedAt": "2026-06-17T09:04:32.087Z",
"npmAuditRan": true,
"pipAuditRan": false,
"promptInjectionRan": true
}open-computer-use is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by Wide-Moat. MCP server that gives any LLM its own computer — managed Docker workspaces with live browser, terminal, code execution, document skills, and autonomous sub-agents. Self-hosted, open-source, pluggable into any model. It has 100 GitHub stars.
open-computer-use failed SkillsLLM's automated security scan, which flagged one or more high-severity issues. Review the Security Report section carefully before using it.
Clone the repository with "git clone https://github.com/Wide-Moat/open-computer-use" and add it to your Claude Code skills directory (see the Installation section above).
open-computer-use is primarily written in Python. It is open-source under Wide-Moat on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh open-computer-use against similar tools.
No comments yet. Be the first to share your thoughts!
Requires a passing catalog security scan. Resolve the flagged issues and resubmit to enable featuring.
MCP server that gives any LLM its own computer — managed Docker workspaces with live browser, terminal, code execution, document skills, and autonomous sub-agents. Self-hosted, open-source, pluggable into any model.
Online demo: chat.yambr.com — Open WebUI with Computer Use already set up, sign in with GitHub or Google. (More ways to try it below.)
See it in action: Demo course on docs.yambr.com — eight live scenarios captured from the chat above (pitch deck, Word doc, Excel, PDF invoice, data chart, live-rendered landing page, web scrape, building a custom skill). Real prompts, real screenshots, copy-pasteable.
If any of this looks useful, a ⭐ on the repo really helps — thanks!

An MCP server that gives any LLM a fully-equipped Ubuntu sandbox with isolated Docker containers. Think of it as your AI's computer — it can do everything a developer can do:
Built for production multi-user deployments. Tested with 1,000+ MAU. Each chat session runs in its own isolated Docker container — the AI can install packages, create files, run servers, and nothing leaks between users. Works seamlessly across MCP clients: start with Open WebUI today, switch to Claude Desktop or n8n tomorrow — same backend, no migration.
| Feature | Open Computer Use | Claude.ai (Claude Code web) | open-terminal | OpenAI Operator |
|---|---|---|---|---|
| Self-hosted | Yes | No | Yes | No |
| Any LLM | Yes (OpenAI-compatible) | Claude only | Any (via Open WebUI) | GPT only |
| Code execution | Full Linux sandbox | Sandbox (Claude Code web) | Sandbox / bare metal | No |
| Live browser | CDP streaming (shared, interactive) | Screenshot-based | No | Screenshot-based |
| Terminal + Claude Code | ttyd + tmux + Claude Code CLI | Claude Code web (built-in) | PTY + WebSocket | N/A |
| Skills system | 13 built-in (auto-injected) + custom | Built-in skills + custom instructions | Open WebUI native (text-only) | N/A |
| Container isolation | Docker (runc), per chat | Docker (gVisor) | Shared container (OS-level users) | N/A |
Works with any MCP-compatible client: Open WebUI, Claude Desktop, LiteLLM, n8n, or your own integration. See docs/COMPARISON.md for a detailed comparison with alternatives.








For all eight live scenarios with prompts you can copy-paste, see the Demo course. See docs/FEATURES.md for architecture details and docs/SCREENSHOTS.md for all screenshots.
Pro tip: Create skills with Claude Code in the terminal, then use them with any model in the chat. Skills are model-agnostic — write once, use everywhere.
Multi-CLI sub-agent runtime (v0.9.2.1+): The sub-agent dispatch supports Claude Code (default), OpenAI Codex, and OpenCode (with OpenRouter / qwen / DeepSeek / 75+ providers). Flip
SUBAGENT_CLI=claude|codex|opencodein.env— see docs/multi-cli.md for the worked OpenCode + qwen3-coder + OpenRouter recipe.
Looking ahead: a Kubernetes-friendly architecture with object-storage-backed user data and squashfs-packaged skills is being designed in docs/future-architecture/. Docker Compose remains the primary supported path.
| Path | URL | What you need | Best for |
|---|---|---|---|
| Free online demo — Open WebUI + Computer Use, models included | chat.yambr.com | GitHub or Google sign-in | Trying it end-to-end in 30 seconds |
| Hosted MCP endpoint — tools only, bring your own LLM | Key at app.yambr.com → connect to https://api.yambr.com/mcp/computer_use |
GitHub/Google sign-in; your own OpenAI / Anthropic / OpenRouter key | Plugging Computer Use into Claude Desktop, n8n, OpenAI Agents SDK |
| Self-host | Quick Start below | Docker, ~15 min first build | Full control, air-gapped, heavy use |
OAuth only — no email/password, no SMS. On chat.yambr.com models are bundled as a free convenience; the hosted API is tools-only. Canonical cloud docs: docs.yambr.com. Repo-side orientation: docs/CLOUD.md.
git clone https://github.com/Wide-Moat/open-computer-use.git
cd open-computer-use
cp .env.example .env
# Edit .env — set OPENAI_API_KEY (or any OpenAI-compatible provider)
# 1. Start Computer Use Server (builds workspace image on first run, ~15 min)
docker compose up --build
# 2. Start Open WebUI (in another terminal)
docker compose -f docker-compose.webui.yml up --build
Open http://localhost:3000 — Open WebUI with Computer Use ready to go.
Note: Two separate docker-compose files:
docker-compose.yml(Computer Use Server) anddocker-compose.webui.yml(Open WebUI). They communicate vialocalhost:8081. This mirrors real deployments where the server and UI run on different hosts.
After adding a model in Open WebUI, go to Model Settings and set:
| Setting | Value | Why |
|---|---|---|
| Function Calling | Native |
Required for Computer Use tools to work |
| Stream Chat Response | On |
Enables real-time output streaming |
Without Function Calling: Native, the model won't invoke Computer Use tools.
| Category | Tools |
|---|---|
| Languages | Python 3.12, Node.js 22, Java 21, Bun |
| Documents | LibreOffice, Pandoc, python-docx, python-pptx, openpyxl |
| pypdf, pdf-lib, reportlab, tabula-py, ghostscript | |
| Images | Pillow, OpenCV, ImageMagick, sharp, librsvg |
| Web | Playwright (Chromium), Mermaid CLI |
| AI | Claude Code CLI, Playwright MCP |
| OCR | Tesseract (configurable languages) |
| Media | FFmpeg |
| Diagrams | Graphviz, Mermaid |
| Dev | TypeScript, tsx, git |
13 built-in public skills + 14 examples:
| Skill | Description |
|---|---|
| pptx | Create/edit PowerPoint presentations with html2pptx |
| docx | Create/edit Word documents with tracked changes |
| xlsx | Create/edit Excel spreadsheets with formulas |
| Create, fill forms, extract, merge PDFs | |
| sub-agent | Delegate complex tasks to Claude Code |
| playwright-cli | Browser automation and web scraping |
| describe-image | Vision API image analysis |
| frontend-design | Build production-grade UIs |
| webapp-testing | Test web applications with Playwright |
| doc-coauthoring | Structur |