From f549a2582745a2b968b5052f2ff3ba24cbbc9570 Mon Sep 17 00:00:00 2001 From: Dillon DuPont Date: Wed, 28 May 2025 16:20:15 -0400 Subject: [PATCH] README updates --- README.md | 41 +++--- README_quickstart_variations.md | 228 -------------------------------- 2 files changed, 16 insertions(+), 253 deletions(-) delete mode 100644 README_quickstart_variations.md diff --git a/README.md b/README.md index 640bf49a..c0bcd410 100644 --- a/README.md +++ b/README.md @@ -21,42 +21,33 @@ # 🚀 Quick Start with a Computer-Use Agent UI -### macOS (Local VMs + Cloud Containers) - -Get started with a Computer-Use Agent UI with a single command: +**Need to automate desktop tasks? Launch the Computer-Use Agent UI with a single command.** +**macOS:** ```bash +# Requires Python 3.11+ /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/scripts/playground.sh)" ``` -This script will: -- Prompt you to choose between [C/ua Cloud Containers](https://trycua.com) or local macOS VMs -- Install Lume CLI for VM management (if needed) -- Pull the latest macOS CUA image (if needed) -- Set up Python environment and install/update required packages -- Launch the Computer-Use Agent UI +
+What does this script do? -**System Requirements:** -- Mac with Apple Silicon (M1/M2/M3/M4 series) -- macOS 15 (Sequoia) or newer -- Disk space for VM images (30GB+ recommended) - -### Windows/Linux (Cloud Containers Only) - -Get started with a Computer-Use Agent UI with a single command: +1. **Asks if you want to use local VMs?** + - **If yes:** + 1. Install VM management CLI + 2. `lume pull macos-sequoia-cua:latest` - Download macOS image + 3. `lume run macos-sequoia-cua:latest` - Start VM +2. `pip install "cua-computer[all]" "cua-agent[all]"` - Install packages +3. `python -m agent.ui.gradio.app` - Launch UI +
+**Windows/Linux:** ```bash +# Requires Python 3.11+ and C/ua Cloud API key pip install "cua-computer[all]" "cua-agent[all]" ; python -m agent.ui.gradio.app ``` -This will: -- Install the required Python packages -- Launch the Computer-Use Agent UI -- Connect to cloud containers through [trycua.com](https://www.trycua.com/) - -**System Requirements:** -- Python 3.11+ -- Internet connection for cloud container access +*How it works: Computer module provides secure desktops (Lume CLI locally, [C/ua Cloud](https://trycua.com) remotely), Agent module handles local/API agents with OpenAI AgentResponse format and [trajectory tracing](https://trycua.com/trajectory-viewer).* ## Supported [Agent Loops](https://github.com/trycua/cua/blob/main/libs/agent/README.md#agent-loops) - [UITARS-1.5](https://github.com/trycua/cua/blob/main/libs/agent/README.md#agent-loops) - Run locally on Apple Silicon with MLX, or use cloud providers diff --git a/README_quickstart_variations.md b/README_quickstart_variations.md deleted file mode 100644 index e8f8cdcd..00000000 --- a/README_quickstart_variations.md +++ /dev/null @@ -1,228 +0,0 @@ -# Quick Start Section Variations - -Here are 5 different variations for the Quick Start section, focused on users who want to use Computer-Use Agent UI: - -## Variation 1: User-Focused with Technical Context - -# 🚀 Quick Start - -**Launch the Computer-Use Agent UI in 60 seconds.** - -## macOS (Local + Cloud) -```bash -/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/scripts/playground.sh)" -``` - -
-What does this script do? - -The playground script automates the complete setup process: - -1. **Install Lume CLI** - ```bash - /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/libs/lume/scripts/install.sh)" - ``` - -2. **Pull the macOS CUA image** - ```bash - lume pull macos-sequoia-cua:latest - ``` - -3. **Run the VM** - ```bash - lume run macos-sequoia-cua:latest - ``` - -4. **Install Python packages** - ```bash - pip install "cua-computer[all]" "cua-agent[all]" - ``` - -5. **Launch the UI** - ```bash - python -m agent.ui.gradio.app - ``` - -You can run these steps manually if you prefer more control over the process. -
- -## Windows/Linux (Cloud) -```bash -pip install "cua-computer[all]" "cua-agent[all]" ; python -m agent.ui.gradio.app -``` - -*The Agent UI uses the Computer module to provide secure macOS/Linux desktops via Lume CLI (local) or [C/ua Cloud](https://trycua.com) (cloud), and the Agent module for local/API agents with OpenAI AgentResponse format and [tracing](https://trycua.com/trajectory-viewer).* - ---- - -## Variation 2: Clear User Intent - -# 🚀 Quick Start - -**Want to use Computer-Use Agents? Get the UI running now.** - -### macOS Users -```bash -/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/scripts/playground.sh)" -``` - -
-What does this script do? - -1. **Install Lume CLI for VM management** - ```bash - /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/libs/lume/scripts/install.sh)" - ``` - -2. **Download the pre-configured macOS image** - ```bash - lume pull macos-sequoia-cua:latest - ``` - -3. **Start the virtual machine** - ```bash - lume run macos-sequoia-cua:latest - ``` - -4. **Install the Python SDK** - ```bash - pip install "cua-computer[all]" "cua-agent[all]" - ``` - -5. **Launch the Computer-Use Agent UI** - ```bash - python -m agent.ui.gradio.app - ``` -
- -### Windows/Linux Users -```bash -pip install "cua-computer[all]" "cua-agent[all]" ; python -m agent.ui.gradio.app -``` - -*Technical details: The UI leverages the Computer module (secure desktops via Lume CLI or [C/ua Cloud](https://trycua.com)) and Agent module (local/API agents with OpenAI AgentResponse format and [tracing](https://trycua.com/trajectory-viewer)).* - ---- - -## Variation 3: Direct and Simple - -# 🚀 Quick Start - -**Get the Computer-Use Agent UI running:** - -```bash -# macOS (local + cloud options) -/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/scripts/playground.sh)" -``` - -
-What does this script do? - -1. `curl -fsSL .../install.sh | bash` - Install Lume CLI -2. `lume pull macos-sequoia-cua:latest` - Download VM image -3. `lume run macos-sequoia-cua:latest` - Start VM -4. `pip install "cua-computer[all]" "cua-agent[all]"` - Install packages -5. `python -m agent.ui.gradio.app` - Launch UI -
- -```bash -# Windows/Linux (cloud containers) -pip install "cua-computer[all]" "cua-agent[all]" ; python -m agent.ui.gradio.app -``` - -*For developers: Uses Computer module (secure desktops via Lume CLI or [C/ua Cloud](https://trycua.com)) + Agent module (local/API agents with OpenAI AgentResponse and [tracing](https://trycua.com/trajectory-viewer)).* - ---- - -## Variation 4: Problem-Solution - -# 🚀 Quick Start - -**Need to automate desktop tasks? Launch the Computer-Use Agent UI.** - -**macOS:** -```bash -/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/scripts/playground.sh)" -``` - -
-What does this script do? - -Behind the scenes, the playground script runs these commands: - -1. **Install Lume CLI** - ```bash - /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/libs/lume/scripts/install.sh)" - ``` - -2. **Pull macOS CUA image** - ```bash - lume pull macos-sequoia-cua:latest - ``` - -3. **Run the virtual machine** - ```bash - lume run macos-sequoia-cua:latest - ``` - -4. **Install Python dependencies** - ```bash - pip install "cua-computer[all]" "cua-agent[all]" - ``` - -5. **Start the Agent UI** - ```bash - python -m agent.ui.gradio.app - ``` -
- -**Windows/Linux:** -```bash -pip install "cua-computer[all]" "cua-agent[all]" ; python -m agent.ui.gradio.app -``` - -*Architecture: Computer module provides secure desktops (Lume CLI locally, [C/ua Cloud](https://trycua.com) remotely), Agent module handles local/API agents with OpenAI AgentResponse format and [tracing](https://trycua.com/trajectory-viewer).* - ---- - -## Variation 5: Ultra Simple - -# 🚀 Quick Start - -**Start using Computer-Use Agents:** - -**macOS:** -```bash -/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/trycua/cua/main/scripts/playground.sh)" -``` - -
-What does this script do? - -1. `lume install` - Install VM management CLI -2. `lume pull macos-sequoia-cua:latest` - Download macOS image -3. `lume run macos-sequoia-cua:latest` - Start VM -4. `pip install "cua-computer[all]" "cua-agent[all]"` - Install packages -5. `python -m agent.ui.gradio.app` - Launch UI -
- -**Windows/Linux:** -```bash -pip install "cua-computer[all]" "cua-agent[all]" ; python -m agent.ui.gradio.app -``` - -*Uses Computer module (secure desktops via Lume CLI or [C/ua Cloud](https://trycua.com)) + Agent module (local/API agents with OpenAI AgentResponse and [tracing](https://trycua.com/trajectory-viewer)).* - ---- - -## Recommendation - -**Variation 1** is the best choice because: - -1. **Clear User Intent**: "Launch the Computer-Use Agent UI" - immediately clear what this does -2. **Time Promise**: "in 60 seconds" sets expectations -3. **Technical Context**: Module details are at the bottom as context, not the focus -4. **User-First**: Focuses on what the user wants to accomplish -5. **Complete**: Still includes all necessary technical information for developers - -This puts the user's goal first while keeping the technical architecture details available for those who need them.