The background
In Twitch's IRC there's a tag called flags, which is obviously undocumented. If Automod detects something in the message the tag will take the form of [START_POS]-[END_POS]:[SCORES]. Where [SCORES] is zero __*__ or more of A.[ASCORE] / I.[ISCORE] / P.[PSCORE] / S.[SSCORE], separated by slashes /. Multiple of these "flags" can be in one message joined by commas, like 1-2:A.1/I.2,3-5:P.6/I.2
Explanation
|Placeholder |Meaning|
|-|-|
| START_POS | Starting position of the word(s) |
| END_POS | Ending position of the word(s) |
| ASCORE | Aggression score of the word(s) |
| ISCORE | Identity language score of the word(s) |
| PSCORE | Profanity score of the word(s) |
| SSCORE | Sexual score of the word(s) |
__*__ Note: Links get no scores, but they get a start and end position.
What should be added?
I would like for these flags to be optionally highlighted over a certain threshold.
Why should it be added?
To aid moderators in detecting words that might be bad depending on the context.
Proof of concept screenshot

How is it currently
At this point Chatterino doesn't show these at all.
Could be a part of moderation mode? I doubt it would be useful to anyone who's not a mod.
Drug substances also get flagged, but they do not get a score. Just like links.
Also, further explanation* for each field:
ISCORESSCOREASCOREPSCORE*Per "How To Use AutoMod" Twitch FAQ article.
Oh, and Twitch user @idmgroup initially made us aware of this undocumented tag.
Chat filters can now be enabled in the web client.

Most helpful comment
Drug substances also get flagged, but they do not get a score. Just like links.
Also, further explanation* for each field:
ISCOREIdentity language - Words referring to race, religion, gender, orientation, disability, or similar. Hate speech falls under this category.
SSCORESexually explicit language - Words or phrases referring to sexual acts, sexual content, and body parts.
ASCOREAggressive language - Hostility towards other people, often associated with bullying.
PSCOREProfanity - Expletives, curse words, and vulgarity. This filter especially helps those who wish to keep their community family-friendly.
*Per "How To Use AutoMod" Twitch FAQ article.
Oh, and Twitch user
@idmgroupinitially made us aware of this undocumented tag.