CO

codex-computer-use

Enables closed-loop GUI automation by interpreting desktop screenshots into discrete action steps.

Install

mkdir -p .claude/skills/codex-computer-use && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/13660" && unzip -o skill.zip -d .claude/skills/codex-computer-use && rm skill.zip

Installs to .claude/skills/codex-computer-use

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Guide Codex CLI to act as a VoiceUse computer-use planner from screenshots. Use when Codex needs to inspect a desktop screenshot and return one safe JSON action for a closed-loop GUI automation task: click, type, key, wait, done, or failed.
240 chars✓ has a “when” trigger
Beginner

Key capabilities

  • Inspect a desktop screenshot
  • Return one JSON object for the next action
  • Choose the smallest safe next action
  • Use 'done' when the task is complete
  • Use 'failed' when the target is not visible

How it works

This skill inspects a desktop screenshot and the current UI state to determine the next action. It returns a single JSON object representing a GUI automation command.

Inputs & outputs

You give it
desktop screenshot, user task, observed region, and recent action history
You get back
one JSON object describing a click, type, key, wait, done, or failed action

When to use codex-computer-use

  • Automate repetitive desktop workflows
  • Navigate complex GUI applications
  • Verify task completion through visual feedback

About this skill

Codex Computer Use

Role

Act as the visual planner for VoiceUse. Inspect the attached screenshot and return exactly one JSON object describing the next local desktop action. VoiceUse executes the action, captures a fresh screenshot, and calls Codex again until the task is done or failed.

Workflow

  1. Read the user task, observed region, and recent action history.
  2. Inspect the screenshot for the current UI state.
  3. Choose the smallest safe next action.
  4. Return one JSON object only. Do not include markdown, prose, or code fences.
  5. Use done only when the screenshot shows the task is complete.
  6. Use failed when the target is not visible, ambiguous, blocked by login/permission, or unsafe to guess.

Action Contract

Load references/action-contract.md for the exact JSON schema, confidence rules, and recovery policy.

Prompt Builder

Use scripts/build_prompt.py when an integration needs deterministic prompt text. It accepts JSON on stdin with:

  • task
  • target_label
  • width
  • height
  • history

It writes the full Codex CLI prompt to stdout.

When not to use it

  • The task requires markdown, prose, or code fences in the output
  • The task is not a closed-loop GUI automation task

Limitations

  • The skill returns only one JSON object per invocation
  • The skill does not include markdown, prose, or code fences in its output
  • The skill uses 'failed' when the target is not visible, ambiguous, blocked by login/permission, or unsafe to guess

How it compares

This skill provides a structured JSON action for GUI automation based on visual inspection, rather than requiring manual interaction or scripting.

Compared to similar skills

codex-computer-use side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
codex-computer-use (this skill)03moReviewBeginner
using-superpowers953moNo flagsBeginner
ultrawork112moNo flagsAdvanced
clawhub253moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

using-superpowers

obra

Use when starting any conversation - establishes mandatory workflows for finding and using skills, including using Skill tool before announcing usage, following brainstorming before coding, and creating TodoWrite todos for checklists

95205

ultrawork

Yeachan-Heo

Parallel execution engine for high-throughput task completion

11184

clawhub

openclaw

Use the ClawHub CLI to search, install, update, and publish agent skills from clawhub.com. Use when you need to fetch new skills on the fly, sync installed skills to latest or a specific version, or publish new/updated skill folders with the npm-installed clawhub CLI.

25151

skill-installer

openai

Install Codex skills into $CODEX_HOME/skills from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).

29141

continuous-learning

affaan-m

Automatically extract reusable patterns from Claude Code sessions and save them as learned skills for future use.

995

memory-keeper-proactive-context-maintenance

b4CU-R4U

Automatically detect and maintain memory freshness by monitoring context staleness, significant code changes, task completions, and phase transitions. Proactively suggests and executes memory sync operations with user confirmation. Use when the user says "sync memory", "update context", or when the Skill detects that context is stale (>2 hours), significant changes have occurred (new commits), tasks completed, or major milestones reached. Replaces passive "context is stale" warnings with active maintenance.

694

Search skills

Search the agent skills registry