Agent Workspace Setup¶
Overview¶
This guide provides minimal steps for bootstrapping the agent development workspace quickly.
Prerequisites¶
- Git installed and configured
- Code editor with Amazon Q plugin (VS Code, Zed, etc.)
- Terminal/command line access
agentctlinstalled (see below)
Install agentctl¶
# macOS/Linux
brew tap geronimo-iia/agent && brew install agentctl
# Windows
choco install agentctl
# or via asdf (version manager)
asdf plugin add agentctl https://github.com/geronimo-iia/asdf-agentctl.git
asdf install agentctl latest
asdf global agentctl latest
# or via cargo
cargo install agent-ctl
Verify: agentctl --version
Installation Process¶
Step 1: Create Workspace Directory¶
# Create workspace directory
mkdir agent-workspace
cd agent-workspace
Step 2: Clone Agent Repositories¶
# Clone core agent repositories
git clone https://github.com/geronimo-iia/agent-foundation.git agent-foundation
git clone https://github.com/geronimo-iia/agent-skills.git agent-skills
git clone https://github.com/geronimo-iia/agent-software.git agent-software
Step 3: Configure IDE Integration¶
Amazon Q Plugin¶
- Install Amazon Q plugin in your IDE
- Open the agent-workspace directory
- Copy rule templates to
.amazonq/rules/for automatic context activation: - From
agent-foundation/rules/templates/(documentation and workspace standards) - From
agent-software/rules/templates/(development standards) - Customize the templates for your workspace structure
- The
.amazonq/rules/directory will automatically apply workspace standards
Step 4: Setup Agent Skills Hub (Optional)¶
# Add the agent-skills hub registry
agentctl hub add --type skills agent-skills https://raw.githubusercontent.com/geronimo-iia/agent-skills/main/index.json --git-url https://github.com/geronimo-iia/agent-skills
# List available skills
agentctl skill list
# Install useful development skills
agentctl skill install agentctl # agentctl management skill
agentctl skill install python-scaffold # Python project scaffolding
agentctl skill install semantic-commit # Semantic commit helpers
Step 5: Ready to Use¶
The workspace is now ready for development:
- agent-foundation/ - Reference stable specifications
- agent-skills/ - Available skills for agent use
- agent-software/ - Software development standards and engineering practices
- agentctl - Validate and generate hub indexes
- Workspace rules - Automatically enforce standards through Amazon Q
For detailed usage, explore the specifications in each repository.