Eliza

Edit the character files

Open src/character.ts to modify the default character. Uncomment and edit.

Custom characters

To load custom characters instead:

• Use pnpm start --characters="path/to/your/character.json"
• Multiple character files can be loaded simultaneously

Add clients

# in character.ts
clients: [Clients.TWITTER, Clients.DISCORD],

# in character.json
clients: ["twitter", "discord"]

Duplicate the .env.example template

cp .env.example .env

* Fill out the .env file with your own values.

Add login credentials and keys to .env

DISCORD_APPLICATION_ID="discord-application-id"
DISCORD_API_TOKEN="discord-api-token"
...
OPENROUTER_API_KEY="sk-xx-xx-xxx"
...
TWITTER_USERNAME="username"
TWITTER_PASSWORD="password"
TWITTER_EMAIL="your@email.com"

Install dependencies and start your agent

pnpm i && pnpm start

Note: this requires node to be at least version 22 when you install packages and run the agent.

Run with Docker

Build and run Docker Compose (For x86_64 architecture)

Edit the docker-compose.yaml file with your environment variables

services:
    eliza:
        environment:
            - OPENROUTER_API_KEY=blahdeeblahblahblah

Run the image

docker compose up

Build the image with Mac M-Series or aarch64

Make sure docker is running.

# The --load flag ensures the built image is available locally
docker buildx build --platform linux/amd64 -t eliza-starter:v1 --load .

Edit the docker-compose-image.yaml file with your environment variables

services:
    eliza:
        environment:
            - OPENROUTER_API_KEY=blahdeeblahblahblah

Run the image

docker compose -f docker-compose-image.yaml up

Set up python environment

Create a new directory for the virtual environment

mkdir ~/python-env

Create a new virtual environment

python3 -m venv ~/python-env/eliza-build

Activate the virtual environment

source ~/python-env/eliza-build/bin/activate

Install setuptools in the virtual environment

pip install setuptools