Change Log



  • Made all commands respond with ‘Usage’ message if the wrong arguments are given
  • Added a link to Model Tuning documentation when model generation fails
  • Made auto filters be at least two characters long


  • Added impersonate command to DeepfakeBot
  • Bug fixes to extract and filter commands


  • Added a message to new users with a link to documentation
  • Training subjects are now linked to trainers - i.e. other users can’t mess with your filters or model settings
  • Changed S3 expiration time to 30 days
  • Added auto filters for extract task, added direct message when extraction task finishes
  • Added newsletter command
  • Added cooldowns to commands
  • Documentation updates: added interacting-with-your-bot page, added notes for Plexi users, added changelog
  • self-deployment branch: Renamed “generate” to “impersonate” for your bot
  • self-deployment branch: Config settings change also updates .json file in S3
  • self-deployment branch: Added catfacts, catpics, and asciify commands


  • Fixed an issue with the extract task related to anmiated emojis by updating version
  • Added a message with a time estimate for extract task
  • Added only a single extract task per user at once


  • Big documentation update
  • Default config changes: removed ‘favorite words’, added owner_id


  • Moved avatar url to private message
  • Fixed an issue with activity plots
  • Added avatar url to deploy self
  • Added user name to message when markov model generator fails


  • Added generate command
  • Made lambda invokations asynchronous
  • Added timestamps to logging
  • Added a stats command
  • Updated readme
  • Documentation setup
  • Changed activity plot to use instead of nick


  • Moved activity plot to lambda function, added a lambda cog base class
  • Extract improvements: added better handling of mentions, renamed channels csv file


  • Removed unique constraint on subjects table
  • Made wordcloud a lambda function, added a lambda layer


  • Release script setup


  • Release script setup


  • First stable release using Markov chains