GitHub - ChrisLamDev/cua-desktop-automation-skills: Router Learning System + macOS desktop automation skills for AI agents using Cua Driver — task routing, fallback chains, self-learning history. · GitHub
/" data-turbo-transient="true" />
Skip to content
Search or jump to...
Search code, repositories, users, issues, pull requests...
-->
Search
Clear
Search syntax tips
Provide feedback
--><br>We read every piece of feedback, and take your input very seriously.
Include my email address so I can be contacted
Cancel
Submit feedback
Saved searches
Use saved searches to filter your results more quickly
-->
Name
Query
To see all available qualifiers, see our documentation.
Cancel
Create saved search
Sign in
/;ref_cta:Sign up;ref_loc:header logged out"}"<br>Sign up
Appearance settings
Resetting focus
You signed in with another tab or window. Reload to refresh your session.<br>You signed out in another tab or window. Reload to refresh your session.<br>You switched accounts on another tab or window. Reload to refresh your session.
Dismiss alert
{{ message }}
ChrisLamDev
cua-desktop-automation-skills
Public
Notifications<br>You must be signed in to change notification settings
Fork
Star
main
BranchesTags
Go to file
CodeOpen more actions menu
Folders and files<br>NameNameLast commit message<br>Last commit date<br>Latest commit
History<br>1 Commit<br>1 Commit
examples
examples
skills/cua-driver-install-macos
skills/cua-driver-install-macos
LICENSE
LICENSE
README.md
README.md
View all files
Repository files navigation
cua-desktop-automation-skills
Router Learning System + macOS desktop automation for AI agents using Cua Driver.
A collection of battle-tested skills and examples for building computer-use agents that control macOS desktops via Cua Driver — with an intelligent task router that learns from past executions.
🌟 Star Feature: Router Learning System
The examples/router.py is a self-improving task routing engine that decides what tool to use based on past success:
Tier 1 — CLI/Git (fastest, no GUI needed)<br>Tier 2 — CDP WebSocket (direct browser DOM manipulation)<br>Tier 3 — Cua Driver (pixel clicks, keystrokes, AX tree)<br>Tier 4 — Vision AI (screenshot analysis for fallback)
How it works
from examples.router import do
# Route + auto-learn — the more you use it, the smarter it gets<br>result = do("git_commit", args={"message": "feat: ..."})<br># → Router checks history, picks best path, records result
Key capabilities:
🧠 Dynamic path sorting — reorders tiers by historical success rate + execution time
📊 SQLite history (~/.hermes/data/router_history.db) — persistent learning across sessions
🔄 Context-aware fallback — auto-escalates through Tier 1 → 2 → 3 → 4 on failure
⏭️ Consecutive failure skip — skips paths that failed 3+ times in a row
🏆 Cold start → warm — static tier order initially, self-optimizes after 5+ samples
Built-in task routes
Task<br>Available paths
create_github_token<br>Device code flow → CDP browser → Safari AX
browser_navigate<br>Cua hotkey (Cmd+L)
browser_fill_input<br>CDP JS → Cua type_text
browser_click_button<br>CDP JS → Cua pixel click
app_hotkey<br>Cua hotkey
app_type<br>Cua type_text
app_click<br>Cua pixel click
app_screenshot<br>screencapture
git_commit<br>git CLI
git_push<br>git CLI
Skills
Skill<br>What it does
cua-driver-install-macos<br>Install, configure, and integrate Cua Driver on macOS — permission granting, MCP setup, Hermes integration
Quick Start
# Clone<br>git clone https://github.com/chrislamlayer1-gif/cua-desktop-automation-skills.git<br>cd cua-desktop-automation-skills
# Try the Router<br>python3 examples/router.py stats<br>python3 examples/router.py git_commit
For Cua Driver Contributors
This repo demonstrates real-world patterns for building computer-use agents with Cua Driver:
Task routing with fallback — the key architectural pattern for reliable desktop automation
Self-learning from history — essential for production agents that improve over time
Multi-tier escalation — CLI → CDP → Cua Driver → Vision — handles all failure modes
Platform Compatibility
macOS (primary target — tested on Apple Silicon)
Requires Cua Driver: brew install trycua/tap/cua
License
MIT — use freely, improve openly.
About
Router Learning System + macOS desktop automation skills for AI agents using Cua Driver — task routing, fallback chains, self-learning history.
Resources
Readme
License
MIT license
Uh oh!
There was an error while loading. Please reload this page.
Activity
Stars
stars
Watchers
watching
Forks
forks
Report repository
Releases
No releases published
Packages
Uh oh!
There was an error while loading. Please reload this page.
Contributors
Uh oh!
There was an error while loading. Please reload this page.
You can’t perform that action at this time.