settings.example.yaml

credentials:
  chatgpt_api_key:        # insert your chatgpt api key here
  # insert your newsapi key here if you want to use the news_api command
  # default is disabled (news_api command). see allowed_commands section below
  newsapi_key:


prompt:
  default_ai_tasks:
    - answering all questions of the humans
    - suggesting useful next steps


  # how much context the history have -> each model can process a different amount of tokens
  # gpt-3.5-turbo around 4000. As this includes the answer also, this value is be a tradeoff
  max_token_len_history: 1500   # 1 token is about 4 characters on average

  # change this if you want your Agent to act as something special (comedian, researcher, ...)
  # describe the role somewhat specific
  ai_default_role: "The Assistant is helpful, friendly and knowledgeable agent The assistant always answers 
  exactly in the format specified. The Assistant does not respond twice with the same command. 
  He wants to make progress."

general:
  model: gpt-3.5-turbo                # see: https://platform.openai.com/docs/models/
  max_response_repairment_attempts: 3 # how often should the model try to repair its own response if broken?
  log_level: info                     # debug < info < warning < error
  key_storage_backend: file     # only file allowed at the moment
  file_storage_backend: file
  query_user_method: cli              # how to ask the user for something? (only cli atm)

  own_names:
    - User

  # what can the bot actually do
  allowed_commands:
    - answer
    - ask_human
    - storage_read
    - storage_write
    - storage_delete
    - read_file
    - write_file
    - list_files
    # - get_datetime
    # - news_api
    - read_website
    - search_web
    - ask_ai_agent
    - read_conversation_history