Chat Message Sent

A chat message was sent in-game, from the Sentinel web interface, or (if enabled) from a linked Discord channel.

Event Data Properties

Property	Type	Notes
source	String	One of Game, Script, Discord, WebService, or WebGroup.
sourceId	String	An identifier of the source. This may be a service id, group id, script id, or Discord channel snowflake.
sourceName	String	A name for the source.
originalMessage	String	The message as it was originally typed. May be censored. Sentinel is not able to capture the uncensored message.
originalLanguage	String	The 2 character language code detected by Sentinel’s translation service. May be xx if translation was not attempted, such as messages sent by the web interface, script, or Discord.
senderName	String	The name of the sender.
scope	String	One of Global, Alliance, Tribe, Local, or Radio.
playerId	String	The Sentinel UUID of the player that sent the message, if the message was sent from the game.
characterId	String	The Sentinel UUID of the survivor that sent the message, if the message was sent from the game.
transations	Object	An object containing translated versions of the original message.
analysis	Object	An object with a guaranteed key of success. When true, a scores key will also be included. See the Score Explanations section below for a detailed explanation of each category.

Sample Event Data

{
  "analysis": {
    "scores": {
      "law": 0.000033,
      "pii": 0.000607,
      "health": 0.000029,
      "sexual": 0.000048,
      "selfharm": 0.000015,
      "financial": 0.000024,
      "violenceAndThreats": 0.005911,
      "hateAndDiscrimination": 0.001099,
      "dangerousAndCriminalContent": 0.008847
    },
    "success": true
  },
  "originalMessage": "I ran out of tranq darts",
  "originalLanguage": "en",
  "translations": {
    "fr": "Ce n'est pas une traduction correcte."
  },
  "playerId": "fc4c921c-ba83-4d1b-8470-a08fedf8246f",
  "characterId": "305b1849-c7ac-5a4b-afe9-86628d91bf23"
}

Score Explanations

Property	Type	Notes
law	Number	Content that contains or tries to elicit detailed or tailored legal advice.
pii	Number	Content that requests, shares, or attempts to elicit personal identifying information such as full names, addresses, phone numbers, social security numbers, or financial account details.
health	Number	Content that contains or tries to elicit detailed or tailored medical advice.
sexual	Number	Material that explicitly depicts, describes, or promotes sexual activities, nudity, or sexual services. This includes pornographic content, graphic descriptions of sexual acts, and solicitation for sexual purposes. Educational or medical content about sexual health presented in a non-explicit, informational context is generally exempted.
selfharm	Number	Content that promotes, instructs, plans, or encourages deliberate self-injury, suicide, eating disorders, or other self-destructive behaviors. This includes detailed methods, glorification, statements of intent, dangerous challenges, and related slang terms
financial	Number	Content that contains or tries to elicit detailed or tailored financial advice.
violenceAndThreads	Number	Content that describes, glorifies, incites, or threatens physical violence against individuals or groups. This includes graphic depictions of injury or death, explicit threats of harm, and instructions for carrying out violent acts. This category covers both targeted threats and general promotion or glorification of violence.
hateAndDiscrimination	Number	Content that expresses prejudice, hostility, or advocates discrimination against individuals or groups based on protected characteristics such as race, ethnicity, religion, gender, sexual orientation, or disability. This includes slurs, dehumanizing language, calls for exclusion or harm targeted at specific groups, and persistent harassment or bullying of individuals based on these characteristics.
dangerousAndCriminalContent	Number	Content that promotes or provides instructions for illegal activities or extremely hazardous behaviors that pose a significant risk of physical harm, death, or legal consequences. This includes guidance on creating weapons or explosives, encouragement of extreme risk-taking behaviors, and promotion of non-violent crimes such as fraud, theft, or drug trafficking.

Notes

The nature of player chat sometimes confuses the translation service. For example, the message u right i guess lol is clearly an English message, but gets detected as Italian. This results in an en key in the translations object that reads you're right, I guess, lol. Do not expect translations to be perfect.

Sentinel currently uses DeepL for its translations and Mistral Moderation for content scoring, but these services could be changed at any time.

Although there is no perfect score to determine whether a chat message belongs to a particular category, 0.5 (50%) is generally considered an acceptable threshold.