Pinned Tweet
CourtAI's free beta is live, and now so is our Discord! 🏀
Inside:
• A direct line to the team. We’re in there daily, reading every piece of feedback and talking props with members
• Extra props and pairings from BigHatProps not released on twitter
• Founding member perks for those who join early
Free to join → discord.gg/GK55YkqHe7
Another example of an easy way to use CourtAI to find potential edges. All I do here is define an outcome: find players that role has changed recently. The agent does the rest by using MCP tools. We can take this even further by asking it to look at lines through CourtAI and find potentially mispriced lines.
There's so much potential to play around with here, and if you notice a data gap that could help you with the way you bet, let us know. We can add it to the MCP and help you and your agent make smarter decisions.
1
7
236
CourtAI retweeted
some cool stuff playing around with the @CourtAI_App mcp
in just a few minutes, Claude learned from all of our twitter plays and found what worked best / makes us profitable.
then spit out a few plays itself using the courtai tools:
give it a read!👇
this merely is the simplest version of the limitless possibilites with the MCP. in today's era of tech and AI, it is crucial to understand how to harness and utilize the raw strength and capacity of such a tool.
with courtai we hope to be the frontier of creating what's next for sports prop research and sports betting.
showing off another cool feature with courtai.app mcp:
give Claude/ChatGPT your favorite capper's twitter plays ➡️ let it automatically learn and find props with his betting style.
cool response from claude on what it took away from @bighatprops plays 👇
give it a shot yourself today for free!
showing off another cool feature with courtai.app mcp:
give Claude/ChatGPT your favorite capper's twitter plays ➡️ let it automatically learn and find props with his betting style.
cool response from claude on what it took away from @bighatprops plays 👇
give it a shot yourself today for free!
some cool plays it was able to find using what it learned from @bighatprops 👇
not official plays or gambling advice**
1
3
162
CourtAI retweeted
Posted some dabble plays that I found using @CourtAI_App on the discord, not going to count them as official plays because they're mostly dabble exclusive, but join if you want to see them. ⬇️
CourtAI retweeted
1/4:
Some interesting scheme nuggets for tomorrows games 🧵:
Researched using @CourtAI_App
Toronto jumps every passing lane: most deflections in the NBA (22.5 per 100), most turnovers forced.
As a result: most free throws allowed per shot.
Drivers vs TOR this year: turnovers +38%, free throws +30% vs their own rate.
Primary ball handlers on the Bucks are Herro, Rollins, Jacquez…
Introducing a new section on CourtAI:
RESEARCH
This is a place to upload reports of any kind, whether it be a prop that you like or a brief that you have automated to run.
You can list it for the public to view with your handle or you can keep it private for yourself.
All cards can be added as graphics to your report, along with your own comments.
Can be created through MCP or on the UI.
Already showed an use case before with my injury brief skill that I have automated to post a report on the site a couple times a day, but also have prop picks that I have on there.
One example usecase of using the MCP to run a long agentic workflow.
Gave Claude a skill as a prompt with the @CourtAI_App MCP to give me an injury brief for tonight.
It worked all 10 games on its own and came back with this. 👇
4
254
CourtAI retweeted
Gave Claude a skill as a prompt with the @CourtAI_App MCP to give me an injury brief for tonight.
It worked all 10 games on its own and came back with this. 👇
NEW:
26-27 season data is loaded in for matchups and game logs, please take it with a grain of salt as we're only a couple games in.
NEW:
Added position groups to targets so you can see how teams perform against broader position groups (GUARDS, WINGS, BIGS).
Positions are heavily fluid in today's NBA and it's really difficult to assign an exact position to a player, but with three broad groups you can find edges in certain team's schemes.
For example, Bigs against Charlotte last year averages 1.3 less REB/36 vs their own average, going under their season rate in 65% of games.
Also, can see our new light mode here for any light mode enjoyers.
showing off what a few of our latest CourtAI.app features can do:
1. Spot a defensive weakness → Turn it into a target.
ask chat/claude to backtest how a play type performed against a team’s scheme, save it, and CourtAI automatically surfaces qualifying players on every slate.
example workflow: 👇
Ask LLM with MCP → Load Agent Response → Track Props
try it for free today at CourtAI.app ⬅️
pretty cool stuff we've been diving into with the CourtAI MCP:
using agents to automate finding advanced trends in NBA teams that would normally take hours
for this case we took a look at how major usage shifts across teams from the off season to now.
give it a read 👇
1/ I ran 2026-27 preseason + opening night through @CourtAI_App:
Last October, big early usage shifts like these kept their direction 92% of the time.
We can try to adjust our assumptions before the books do.
Usage and shot-mix changes worth watching 🧵
4
210
Want to start a short series showing off what the MCP allows you to do with an agent like Claude or ChatGPT.
1) Demonstrating how we quickly identify a potential weakness to exploit against the Kings. Would really recommend watching the video through and trying it out.
introducing CourtAI
the worlds first agent based sports prop research tool, now with LLM MCP connection.
give it a try ⬇️
CourtAI.app
10
908



















