Duration	30 minutes
Module	M1.1 - Introduction to Voice AI

Objectives

• Install the SignalWire Agents SDK
• Configure development environment
• Verify installation with CLI tools

Prerequisites

• Python 3.10+ installed
• Terminal/command line access
• Text editor or IDE (VS Code recommended)

---

Exercise 1: Create Working Directory

Create a dedicated directory for training:

mkdir -p ~/signalwire-training
cd ~/signalwire-training

Checkpoint: You should be in the signalwire-training directory.

---

Exercise 2: Create Virtual Environment

Set up an isolated Python environment:

# Create virtual environment
python3 -m venv .venv

# Activate it
source .venv/bin/activate  # macOS/Linux
# .venv\Scripts\activate   # Windows

Checkpoint: Your prompt should show (.venv) prefix.

---

Exercise 3: Install SDK

Install the SignalWire Agents SDK:

pip install signalwire-agents

Wait for installation to complete.

Checkpoint: No errors in installation output.

---

Exercise 4: Verify Core Import

Test that the SDK loads correctly:

python -c "from signalwire_agents import AgentBase; print('SUCCESS: AgentBase imported')"

Expected Output:

SUCCESS: AgentBase imported

---

Exercise 5: Verify CLI Tools

Test the command-line tools:

# Test swaig-test
swaig-test --help

# Test sw-agent-init
sw-agent-init --help

Checkpoint: Both commands show help text without errors.

---

Exercise 6: Create Environment File (Optional)

Environment files (.env) store configuration. To use them, first install python-dotenv:

pip install python-dotenv

Create a .env file for configuration:

cat > .env << 'EOF'
# SignalWire Credentials (fill in later)
SIGNALWIRE_SPACE_NAME=
SIGNALWIRE_PROJECT_ID=
SIGNALWIRE_TOKEN=

# Agent Authentication (SDK reads these automatically when loaded)
SWML_BASIC_AUTH_USER=signalwire
SWML_BASIC_AUTH_PASSWORD=training123
EOF

Note: The .env file is not loaded automatically. You must add from dotenv import load_dotenv and call load_dotenv() at the start of your agent script. We’ll cover this in Module 1.4.

Checkpoint: File .env exists with template content.

---

Exercise 7: Create Test Agent

Create a simple test agent:

cat > test_agent.py << 'EOF'
#!/usr/bin/env python3
"""Test agent for environment verification."""

from signalwire_agents import AgentBase

agent = AgentBase(name="test-agent")

agent.prompt_add_section("Role", "You are a test agent.")

if __name__ == "__main__":
    print("Environment setup successful!")
    print("Agent name:", agent.name)
EOF

Run it:

python test_agent.py

Expected Output:

Environment setup successful!
Agent name: test-agent

---

Exercise 8: Test SWML Generation

Use swaig-test to verify agent output:

swaig-test test_agent.py --dump-swml

Expected Output: JSON document with SWML structure including your prompt.

---

Verification Checklist

• Virtual environment created and activated
• SDK installed without errors
• AgentBase imports successfully
• swaig-test --help works
• sw-agent-init --help works
• .env file created
• Test agent runs
• SWML generation works

---

Troubleshooting

“python: command not found”

Try python3 instead of python.

“No module named signalwire_agents”

Ensure virtual environment is activated (you see (.venv) in prompt).

“Permission denied”

On Unix, you may need: chmod +x test_agent.py

swaig-test not found

Ensure you’re in the activated virtual environment where SDK was installed.

---

Summary

You have successfully:

1. Created a dedicated training directory
2. Set up an isolated Python environment
3. Installed the SignalWire Agents SDK
4. Verified all components work correctly

Next: Module 1.2 - Your First Agent