Twitter Client

The Twitter Client enables your agent to interact through Twitter, supporting both group chats and direct messages. Visit Elizaos Twitter/x client docs for more details.

Installation

1. Install required packages:

   npm

   npm i github:elizaos-plugins/client-twitter @elizaos/[email protected]

   pnpm

   pnpm add github:elizaos-plugins/client-twitter @elizaos/[email protected]

   yarn

   yarn add github:elizaos-plugins/client-twitter @elizaos/[email protected]

2. Configure environment variables:

    # Twitter API Credentials
    TWITTER_USERNAME=           # Your Twitter/X username
    TWITTER_PASSWORD=           # Your Twitter/X password
    TWITTER_EMAIL=              # Your Twitter/X email
    TWITTER_2FA_SECRET=         # Optional: 2FA secret for login
   
    # Twitter Client Configuration
    TWITTER_DRY_RUN=false      # Set to true for testing without posting
    MAX_TWEET_LENGTH=280       # Default tweet length limit
    TWITTER_SEARCH_ENABLE=false # Enable search functionality
    TWITTER_RETRY_LIMIT=5      # Login retry attempts
    TWITTER_POLL_INTERVAL=120  # Poll interval in seconds
    TWITTER_TARGET_USERS=      # Comma-separated list of target users
   
    # Post Generation Settings
    TWITTER_ENABLE_POST_GENERATION=true
    POST_INTERVAL_MIN=90       # Minimum interval between posts (minutes)
    POST_INTERVAL_MAX=180      # Maximum interval between posts (minutes)
    POST_IMMEDIATELY=false     # Skip approval workflow
   
    # Action Processing
    ENABLE_ACTION_PROCESSING=false
    ACTION_INTERVAL=5          # Action check interval (minutes)
    MAX_ACTIONS_PROCESSING=1   # Maximum concurrent actions
   
    # Spaces Configuration (Optional)
    TWITTER_SPACES_ENABLE=false
    ELEVENLABS_XI_API_KEY=     # Required for TTS in Spaces
   
    # Approval Workflow (Optional)
    TWITTER_APPROVAL_DISCORD_BOT_TOKEN=
    TWITTER_APPROVAL_DISCORD_CHANNEL_ID=
    TWITTER_APPROVAL_CHECK_INTERVAL=300000  # 5 minutes in milliseconds

Basic Setup

import { AgentBuilder, ModelProviderName } from "@iqai/agent";
import TwitterClient from "@elizaos/client-twitter";
import SqliteAdapter from "@elizaos/adapter-sqlite";


async function main() {
  const agent = new AgentBuilder()
    .withDatabase(SqliteAdapter)
    .withClient(TwitterClient)
    .withModelProvider(
      ModelProviderName.OPENAI,
      process.env.OPENAI_API_KEY
    )
    .withCharacter({
      name: "TwitterBot",
      bio: "A helpful Twitter assistant",
      username: "brain_bot"
    })
    .build();

  await agent.start();
}

main().catch(console.error);