Quick Start
Get up and running with ID Agents in minutes
Quick Start#
This guide walks you through installing ID Agents, launching the interactive CLI, deploying your first agents, and sending them messages.
Prerequisites#
Before you begin, make sure you have:
- Node.js 20 or later
- PostgreSQL (for agent state persistence)
- Claude Code or an Anthropic API key
- id-cli (optional, for onchain agent registration)
Setup#
Clone the repository and install dependencies:
git clone https://github.com/idchain-world/id-agents.git
cd id-agents
npm install
Copy the example environment file and configure it:
cp env.example .env
Edit .env and set at minimum the DATABASE_URL pointing to your PostgreSQL instance:
DATABASE_URL=postgresql://user:password@localhost:5432/id_agents
If you are on a Claude Max plan, run claude login first -- no API key is needed. Otherwise, set ANTHROPIC_API_KEY in your .env file.
Running the Interactive CLI#
Start the manager and CLI with:
npm run id-agents
The manager launches on port 4100 by default. To use a custom port:
npm run id-agents -- --port 5000
Or set it via environment variable:
MANAGER_PORT=5000 npm run id-agents
When the manager starts, agents will be assigned ports starting at 4101 (or your custom port + 1).
Deploying Agents#
Use the /deploy command with a YAML configuration file to spin up agents:
/deploy configs/my-team.yaml
A minimal YAML config looks like this:
version: "1"
team: my-team
defaults:
runtime: claude-code-cli
model: claude-opus-4-6
agents:
- name: coder
description: "Writes and reviews code"
workingDirectory: /path/to/project
- name: researcher
description: "Research and analysis"
workingDirectory: /path/to/research
Once deployed, each agent runs as its own local process with a dedicated port and REST-AP endpoints.
Talking to Agents#
Send a message to any deployed agent:
/ask coder Write a hello world function in TypeScript
The agent processes your message asynchronously. You can check for responses:
/news coder
Use /news -l coder to see full message content.
Broadcast to All Agents#
Send a message to every agent at once:
/ask * What is your current status?
Clear a Session#
Start a fresh conversation with an agent:
/clear coder
Registering Agents Onchain#
If you have id-cli installed and ID_REGISTRAR_PRIVATE_KEY set in your .env, you can register an agent on ID Chain:
/register coder
This assigns the agent a permanent ENS-based identity like x.agent-15.sep.xid.eth.
Using the Remote API#
The manager exposes a /remote endpoint for programmatic access. You can execute any CLI command via HTTP:
curl -X POST http://localhost:4100/remote \
-H "Content-Type: application/json" \
-d '{"command":"/agents"}'
This is useful for integrating ID Agents into automation workflows and external applications.
Next Steps#
- CLI Reference -- Full command list and usage
- Configuration -- All YAML and environment options
- REST-AP Protocol -- How agents discover and communicate
- Skills -- Extend what agents can do