Keep track of every change to the Bodyguard.ai API.

All updates and additions to the Bodyguard.ai API are listed in chronological order in this changelog.

April 17, 2023

We are excited to announce a new feature that we believe will have a significant impact on your community: automatic author decision making. Much like message moderation, this feature allows you to set up automatic actions that will be taken on authors based on their behavior within your communities. For example, you can easily ban scammers or other authors who exhibit highly toxic behavior. To take advantage of this feature, you will need to subscribe to our webhooks with AUTHOR_UPDATED events.

March 13, 2023

  • Full re-design of our developer documentation website 💪
  • Improved taxonomy detection:
  • Type
    • Rename SUPPORTIVE to POSITIVE
    • Rename NOISE to UNDESIRABLE
    • Add HATE_SPEECH
    • Add CRITICISM
  • Directed at
    • Add ENTITY
  • Severity
    • Rename VERY_HIGH to CRITICAL
  • Classification
    • Rename HOMOPHOBIA to LGBTQIAPLUS_PHOBIA
    • Add FLOOD
    • Add SELF_HARM
    • Add DOXXING
    • Add VULGARITY
    • Add SEXUALLY_EXPLICIT
    • Add DRUG_EXPLICIT
    • Add WEAPON_EXPLICIT
    • Add PII
    • Add UNDERAGE_USER
    • Add GEOPOLITICAL
    • Add FAIRPLAY

December 30, 2022

  • Add detection for classification MISOGYNY.
  • Set param publishedAt as mandatory to the analyze endpoint.

September 29, 2022

August 24, 2021

  • Say hello to webhooks, which allow you to create advanced workflows based on Bodyguard.ai actions.
  • Bodyguard.ai can identify spam messages using asynchronous analysis. Combined with our webhooks, you'll receive notifications for every identified spam message.
  • Rename context.content to context.post in the analyze endpoint.

June 28, 2021

March 26, 2021

  • Add defaultLanguage field that behave as the current language field.
  • The language field now force the language detection if set.

September 17, 2021

  • Here comes our new documentation for Bodyguard.ai! 💪
  • We have designed this API to be simple and straightforward, with a unique /analyze endpoint.