Agent skill

Cua Driver

by bastani-inc in bastani-inc/atomic

Drive a native GUI app (macOS, Windows, Linux) via the cua-driver CLI (default) or MCP server; snapshot its accessibility tree, act through snapshot-bound element tokens, native menu paths, exact…

Custom licenceAuto-check passedFrontend & Design

Install Cua Driver

skills CLI
$ npx skills add bastani-inc/atomic --skill cua-driver -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install bastani-inc/atomic cua-driver --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/bastani-inc/atomic.git skills-src && mkdir -p .claude/skills && cp -r skills-src/packages/subagents/skills/cua-driver .claude/skills/cua-driver && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
cua-driver
GitHub stars
855
Token cost
~2k tokens
SKILL.md length
691 words
Files
11
Skills in repo
15
Repo updated
First seen
Licence
Custom licence

At a glance

Drive a native GUI app (macOS, Windows, Linux) via the cua-driver CLI (default) or MCP server; snapshot its accessibility tree, act through snapshot-bound element tokens, native menu paths, exact…

  • Works in 7 steps: Select the exact target on each action.… → Observe before input and verify after… → Use returned tokens, never invented… → …
  • The user asks you to operate
  • SKILL.md covers Act, Detect, Rules and Failure map, plus 2 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Cua Driver is an agent skill from bastani-inc/atomic. Drive a native GUI app (macOS, Windows, Linux) via the cua-driver CLI (default) or MCP server; snapshot its accessibility tree, act through snapshot-bound element tokens, native menu paths, exact window geometry, or pixel coordinates, and verify from fresh state. Use when the user asks you to operate, drive, automate, or perform a GUI task in a real application on the host, or to continue, resume, or recall recent Cua activity.

Its SKILL.md is about 2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 10 other files (for example `BROWSER.md`, `EMBEDDING.md` and `LINUX.md`).

It sits in Frontend & Design. It works with Linux, Model Context Protocol and macOS. The repository describes itself as: The verifiable coding agent runtime. Define your coding agent's process in natural language with stages, checks, and approval gates instead of hoping it follows your…

When your agent uses it

  • The user asks you to operate
  • Perform a GUI task in a real application on the host
  • Recall recent Cua activity

Example prompts

  • “/cua-driver”

Requirements

  • Pre-approved tools (allowed-tools): Bash(cua-driver:*)

Workflow steps

7 steps, taken from the first numbered list in SKILL.md.

  1. Select the exact target on each action. A session is lifecycle metadata, not capture scope or permission authority.
  2. Observe before input and verify after it. effect:"unverifiable" and a successful exit are not task success; never replay a partial…
  3. Use returned tokens, never invented indices. A fresh snapshot replaces prior element handles and lists them in invalidated_snapshot_ids…
  4. Keep background window actions non-interfering. Foreground delivery and desktop input require authorization for visible control; an…
  5. Never infer pixels from a missing image, a different window, or an unaccounted-for resized preview. Capture failure and an empty…
  6. Keep one controller for a shared desktop. Distinct sessions/cursors do not isolate focus, keyboard input, application state, or snapshot…
  7. User/system permission prompts belong to the user or trusted host. Never alter browser profiles or security settings as hidden setup…

What it can do on your machine

Read from SKILL.md and the folder at commit 23dbfd7. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Bash(cua-driver:*)

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Cua Driver loads about 2k tokens when it runs. Until then it costs about 111 tokens; SKILL.md has 691 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~111
When it runs · the whole SKILL.md, loaded when a task matches
~2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

Its licence (Custom licence) doesn't allow us to republish the file, so here is its outline and opening line. It has 691 words (~2,040 tokens).

“Operate one exact target, observe its state, act once, and verify the user's postcondition.”

— opening of SKILL.md by bastani-inc, Custom licence
name
cua-driver
allowed-tools
Bash(cua-driver:*)
hidden
true
metadata.github-path
libs/cua-driver/rust/Skills/cua-driver
metadata.github-ref
refs/tags/cua-driver-rs-v0.34.0
metadata.github-repo
https://github.com/trycua/cua
metadata.github-tree-sha
bc6d2b59a453e10321e30e58203294fee9d4bc28
version
0.34.0

Read the full SKILL.md on GitHub

Files

SKILL.md and 10 other files in packages/subagents/skills/cua-driver of bastani-inc/atomic.

  • SKILL.md
  • BROWSER.md
  • EMBEDDING.md
  • LINUX.md
  • MACOS.md
  • README.md
  • RECORDING.md
  • RUNTIME.md
  • VISUAL.md
  • WINDOWS.md
  • WORKFLOW.md

Open the folder on GitHubat commit 23dbfd7

Compare with similar skills

Cua Driver next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Cua Driver compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Cua Driver this skillbastani-inc/atomic855—~2kAutomated safety check: PassCustom licence
Gearcoleco Debuggingdrhelius/Gearcoleco142—~3.5kAutomated safety check: PassGPL-3.0
Jarvis Setupethanplusai/jarvis846—~2.5kAutomated safety check: NotesCustom licence
Cortex Mem MCPsopaco/cortex-mem313—~2.8kAutomated safety check: PassMIT
Mps Project ManagementJetBrains/MPS1.7k—~2.2kAutomated safety check: PassApache-2.0
Gearcoleco Romhackingdrhelius/Gearcoleco142—~3.9kAutomated safety check: PassGPL-3.0

Similar skills

  • Gearcoleco Debugging

    drhelius/Gearcoleco

    Debug and trace ColecoVision and Super Game Module games using the Gearcoleco emulator MCP server.

    142 GitHub stars~3.5k tokensUpdated yesterday
    DevelopmentAuto-check passed
  • Jarvis Setup

    ethanplusai/jarvis

    A skill your agent uses when helping someone install, configure, or debug a fresh clone of JARVIS (this repo) — especially "the mic doesn't work", "JARVIS says his language systems are down", any…

    846 GitHub stars~2.5k tokensUpdated 1 mo ago
    Frontend & DesignAuto-check: notes
  • Cortex Mem MCP

    sopaco/cortex-mem

    Persistent memory enhancement for AI agents. An agent skill from sopaco/cortex-mem.

    313 GitHub stars~2.8k tokensUpdated 2 mo ago
    Agent WorkflowsAuto-check passed
  • Official

    Open an MPS project in a running or freshly started MPS instance when MCP tools fail because no project is open (welcome screen), close an open project with mpsmcpcloseproject, or create a new empty…

    1.7k GitHub stars~2.2k tokensUpdated yesterday
    Agent WorkflowsAuto-check passed
  • Gearcoleco Romhacking

    drhelius/Gearcoleco

    Hack, modify, and translate ColecoVision and Super Game Module ROMs using the Gearcoleco emulator MCP server.

    142 GitHub stars~3.9k tokensUpdated yesterday
    Agent WorkflowsAuto-check passed
  • Healthmd CLI

    CodyBontecou/health-md

    Safely install and use the Health.md CLI and MCP server to query user-authorized health data, chart typed metrics, inspect sleep and workouts, export scoped or complete public/authorized Apple…

    230 GitHub stars~4.8k tokensUpdated yesterday
    Productivity & AutomationAuto-check passed

More from bastani-inc/atomic

All 15 skills in this repo
  • How

    bastani-inc/atomic

    A skill your agent uses for "how does X work", code walkthroughs before changing something, and placement / ownership / layering questions ("where should this live", "which package owns this", "is…

    855 GitHub starsUsed in 3 repos~2.3k tokens
    Auto-check passed
  • Liteparse

    bastani-inc/atomic

    A skill your agent uses whenever a task involves a document file (PDF, DOCX, PPTX, XLSX, or image) and you need to read it or pull text, tables, or specific values out of it — to answer a question…

    855 GitHub stars~1.4k tokensUpdated today
    Auto-check passed
  • Bun

    bastani-inc/atomic

    A skill your agent uses when building, testing, and deploying JavaScript/TypeScript applications.

    855 GitHub stars~2.8k tokensUpdated today
    Auto-check passed
  • Feedback

    bastani-inc/atomic

    Draft, revise and post privacy-scrubbed Atomic bug reports or enhancements through ordinary conversation.

    855 GitHub stars~1.1k tokensUpdated today
    Auto-check passed
  • Prek

    bastani-inc/atomic

    A skill your agent uses when setting up or running hooks with prek in any repository.

    855 GitHub stars~2.1k tokensUpdated today
    Auto-check passed
  • TDD

    bastani-inc/atomic

    Test-driven development with red-green-refactor loop, plus a value gate and audit workflow for tests.

    855 GitHub stars~1.4k tokensUpdated today
    Auto-check passed

Questions about Cua Driver

What does Cua Driver do?

Drive a native GUI app (macOS, Windows, Linux) via the cua-driver CLI (default) or MCP server; snapshot its accessibility tree, act through snapshot-bound element tokens, native menu paths, exact…. Cua Driver is an agent skill from bastani-inc/atomic. Drive a native GUI app (macOS, Windows, Linux) via the cua-driver CLI (default) or MCP server; snapshot its accessibility tree, act through snapshot-bound element tokens, native menu paths, exact window geometry, or pixel coordinates, and verify from fresh state.

When should I use Cua Driver?

Cua Driver fits situations like: the user asks you to operate; perform a GUI task in a real application on the host; recall recent Cua activity.

How do I install Cua Driver in Claude Code?

Run `npx skills add bastani-inc/atomic --skill cua-driver -a claude-code`. Or copy the skill folder (packages/subagents/skills/cua-driver in bastani-inc/atomic) into .claude/skills/cua-driver in your project. Claude Code loads it when a task matches its description.

How do I install Cua Driver in Codex?

Run `npx skills add bastani-inc/atomic --skill cua-driver -a codex`. Or copy the skill folder (packages/subagents/skills/cua-driver in bastani-inc/atomic) into .agents/skills/cua-driver in your project. Codex loads it when a task matches its description.

Can I use Cua Driver in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add bastani-inc/atomic --skill cua-driver -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cua-driver, .gemini/skills/cua-driver, .github/skills/cua-driver and .opencode/skills/cua-driver in your project.

What does Cua Driver need to run?

SKILL.md names no scripts, command-line tools or credentials: Cua Driver is instructions for the agent only. Its frontmatter pre-approves these tools: Bash(cua-driver:*).

Does Cua Driver access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Cua Driver safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Cua Driver use?

Cua Driver has a licence file (the repository's licence) that doesn't match a standard licence. Read it on GitHub before reusing the skill.

How many tokens does Cua Driver use?

About 2k tokens (SKILL.md is roughly 8.2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Cua Driver?

Skills that share tags, products or a category with Cua Driver: Gearcoleco Debugging (drhelius/Gearcoleco, 142 stars), Jarvis Setup (ethanplusai/jarvis, 846 stars), Cortex Mem MCP (sopaco/cortex-mem, 313 stars) and Mps Project Management (JetBrains/MPS, 1.7k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Cua Driver?

bastani-inc (a GitHub organization) maintains it in bastani-inc/atomic, which has 855 GitHub stars. The repository holds 15 skills in this directory. The repository was last updated on October 10, 2026.

Source: bastani-inc/atomic on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.