Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add enwong93-sketch/devspace-ultra --skill computer-usegit clone --depth 1 https://github.com/enwong93-sketch/devspace-ultraWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/enwong93-sketch/devspace-ultra/computer-use)<a href="https://agentmods.dev/skills/enwong93-sketch/devspace-ultra/computer-use"><img src="https://agentmods.dev/badge/skills/enwong93-sketch/devspace-ultra/computer-use/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/enwong93-sketch/devspace-ultra/computer-use"><img src="https://agentmods.dev/badge/skills/enwong93-sketch/devspace-ultra/computer-use.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00028 | $0.00495 |
| Opus 5.5 | $0.00011 | $0.00198 |
| Sonnet 5.5 | $0.00006 | $0.00099 |
| Haiku 4.5 | $0.00003 | $0.00049 |
Grade A, and why
codex-computer-use scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 22d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Codex Computer Use routing
Use codex_computer_use automatically when a task requires observing or operating a Windows graphical application, including an ordinary Chrome or Edge browser window: visible windows, buttons, menus, dialogs, screenshots, accessibility state, clicking, typing, scrolling, dragging, navigation through the visible address bar, or rendered-state verification.
Do not use it for repository/file edits, shell commands, APIs, database queries, or source-code search. DevSpace's obsolete custom Chrome-extension driver has been removed; do not recreate, pair, claim, or route browser work through a second implementation.
The DevSpace tool is only a structured adapter over the installed OpenAI bundled Computer Use runtime. The execution path is the shared persistent Codex node_repl importing @oai/sky; DevSpace does not implement SendInput, UI Automation, screenshot capture, Chrome Debugger/CDP control, Selenium, Playwright, PowerShell UI automation, or another GUI/browser driver.
Follow the official observe/action discipline:
- Call
list_appsorlist_windowsand choose exactly one returned target window. - Call
get_window_stateto observe. Use accessibility indexes only from that observation; use screenshot IDs only from that observation. - Perform at most one state-changing action.
- Immediately call
get_window_stateagain before deciding the next action. - Never reuse stale indexes, coordinates, or screenshot IDs after the UI changes.
- When Computer Use is finished, make one final read-only observation with
input.release_control=true. This explicit release is mandatory before the Agent's visible final response or before switching to unrelated non-Computer-Use work; it clears the blue takeover state immediately. If a Computer Use operation fails, DevSpace releases the takeover state automatically as a fail-safe.
ChatGPT Classic desktop windows may be controlled through native Computer Use when the user explicitly requests that app in the current task. First verify the exact target window and conversation, and keep actions scoped to that target. Do not automate terminal applications, Windows Run, authentication/password/security/privacy UI, or the Codex app UI. Higher-priority assistant safety rules and action-time confirmation requirements still apply even though DevSpace's local execution policy is full-access/approval-never.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 22d ago Changed ed6984c6ef81
- 1mo ago Changed · +1 lines 7e596e836737
- 1mo ago Changed 363e9e3915a7
- 1mo ago First seen · 23 lines · 28 tokens per session scan A cc0f39ad2fae
codex-computer-use is a skill published in the GitHub repository enwong93-sketch/devspace-ultra (58 stars, last pushed 16d ago), licensed MIT. It adds 28 tokens to every session and 495 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-10.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…