LogoLogo
  • πŸŽ‰Welcome To Fabelis.AI
  • ❓What Is Fabelis?
  • πŸ›£οΈRoadmap
  • 🀝How To Contribute
  • Playground
    • πŸ›What is the Playground?
    • Chatroom
    • Tool Configuration
    • Character
    • Settings
  • MCP
    • Overview
    • Run a Client
  • Create a Server
  • Active Backends
  • AGENT
    • πŸƒHow To Run
    • 🏌️Character File
    • CLI Chatbot
    • Twitter Agent
    • πŸ‡ΊπŸ‡ΈTruth Social
    • Discord Bot
    • Telegram Bot
  • What We Support
  • Environment
  • Support
Powered by GitBook
On this page
  • Environment
  • Config
  • Interaction
  1. AGENT

Twitter Agent

Here's an example showing you how to setup the .env, config, and run the Twitter client.

PreviousCLI ChatbotNextTruth Social

Last updated 5 months ago

Environment

Set these .env variables by visiting your . Haven't set this up? Learn more .

Name

TWITTER_API_KEY

TWITTER_API_SECRET

TWITTER_ACCESS_TOKEN

TWITTER_ACCESS_TOKEN_SECRET

You MUST regen tokens after changing this setting.

Config

Twitter client be ran with other clients. These other clients are ran in concurrency.

Completion provider can be ANY provider you'd like, and "embedding_provider" "db" ARE REQUIRED.

{
    "clients": {
        "twitter": {
            // Set the min/max range for delay between posts in MINUTES
            "post_delay": [
                10,
                20
            ],
            // Set the min/max range for delay between replies in MINUTES
            "reply_delay": [
                10,
                20
            ],
            // Set delay in MINUTES after searching twitter for relevant posts when creating post response
            "search_delay": 1,
            // Set HARD delay in MINUTES between actions
            "delay": 0,
            // If TRUE agent wont post/reply live but will print its responses in the console
            "debug": true
        }
    },
    "completion_provider": "anthropic",
    "embedding_provider": "local",
    "db": "local"
}

Interaction

No interaction is needed on your end once its running. The agent will print logs related to actions taken, posts fetched, and sleep durations between actions.

Twitter developer dashboard
here