Quickstart¶
Go from installation to your first message in 5 minutes.
Step 1: Install¶
Verify the installation:
Virtual Environment
It's recommended to install in a virtual environment to avoid dependency conflicts:
Step 2: Run the Setup Wizard¶
The wizard guides you through:
- Choose a model provider — OpenAI, Anthropic, Gemini, Bedrock, OpenRouter, or OpenAI-compatible endpoints
- Enter your API key — the key for your chosen provider
- Select a model — e.g.,
gpt-4o,claude-sonnet-4-20250514,gemini-2.0-flash - Basic settings — Agent name, language preference, etc.
Configuration is saved to ~/.echo-agent/config.yaml.
Manual Configuration
Skip the wizard and edit the config file directly:
Step 3: Start the Agent¶
On successful startup, you'll see output like:
[INFO] Echo Agent v0.3.8 starting...
[INFO] Model: openai/gpt-4o
[INFO] Memory: loaded (42 entries)
[INFO] Skills: 12 active
[INFO] Channels: cli
[INFO] Ready. Type your message below.
Step 4: Send Your First Message¶
Type a message directly in the terminal and press Enter:
The Agent will respond and remember the conversation. You can continue chatting — it maintains context across turns.
Step 5: Verify Success¶
Confirm that everything is working:
# Check running status
echo-agent status
# View cost statistics
echo-agent cost
# List loaded skills
echo-agent skill list
Verify Memory
Restart the Agent and ask what you talked about last time. If it recalls, the memory system is working correctly.
Next Steps¶
You've successfully run Echo Agent. Here's where to go next:
- Connect more platforms — Integrate with DingTalk, WeChat, Slack, and more. See Channel Configuration
- Run in background — Configure as a system service for 24/7 availability. See Deployment Guide
- Open the Dashboard — Manage your Agent via the web panel:
- Explore skills — View and manage the Agent's skill library:
- Scheduled tasks — Have the Agent run tasks on a schedule: