gorilli retweeted
scanned ~2,000 websites with LLM Scan so far
median score: 51/100
more than 1 in 4 score under 40
only ~2% score 85+
and these are the people who cared enough to run a scan
most sites are still not ready for AI crawlers
gorilli retweeted
hot take: llms.txt won't get you cited
Google (June docs update): Search ignores it, AI features included
Ahrefs, 137k sites: 97% of llms.txt files got zero requests in May
it's a file, not a strategy
what matters:
- can AI crawlers read you
- does AI actually mention you
Gmonad to you Anon, yes say it back!
"Become a person know for figuring it out"
~~Keone Hon 2026
Have you figured it out yet ?
Check out signals now to figure out what your favorite pair on monad is doing 👇
Gorillionai.re
Stay moisturized and happy!
Above all, have a good week!
Have a good week Nads!
Become a person known for “Figuring It Out” no matter what. Incredibly valuable and evergreen
Your current trials are opportunities to hone the blade!
1
4
279
All assets signals will be live on Gorillionaire soon. Super fast signal Engine for Super performant chain, the storm is coming anon.
Believe in something 🟣
The monad client is super performant, has full parity with Eth interfaces, supports large valsets, and runs on minimal hardware
From that baseline, pushing forward
1
149
Gmonad to the few Nads left, zoom in and out, it doesn't matter, what do you see ?
I see generational bottoms 🙂↔️
Be on time on your trades with signals 👇
gorillionai.re
What did you do last week? I'm on a shipping spree over here, Focus Mode never went off!
🦍 Gorillionaire – Weekly Update
Week 3: January 12 – January 18, 2026
Summary
This week was all about getting deeper into the prediction engine. We kept grinding on the prototype that transforms our signals into actual price forecasts, the math is getting interesting but also messier than expected. Also started exploring a potential partnership that could change how we think about signal distribution.
Highlights
- Prediction Model Deep Dive: Continued building the prototype that uses our signals and tracked wallet data to predict token prices. Been testing different time horizons, 4-hour predictions show 68% accuracy, but 24-hour drops to 52%. The challenge isn't the signals themselves, it's handling the noise from low-volume Monad tokens where one $50k trade can move price 15%. Currently experimenting with different types of confidence scores.
- Data Pipeline Optimization: Had to rebuild how we feed historical data into the model. Original approach was pulling everything fresh each run, which took 12+ seconds. Switched to incremental updates with caching, now predictions generate in under 3 seconds. Small win but feels way better for users.
- Partnership Exploration: Started conversations with a potential partner who's interested in our signal infrastructure. Can't share details yet, but it's the kind of integration that could put our signals in front of 10x more users. Still early discussions, but the technical integration looks feasible.
In Progress
- Prediction Algorithm Refinement: The core model works, but needs smarter handling of market conditions. Working on conditional logic that adjusts the model based onthe overall Monad ecosystem volatility. Probably need another week of backtesting.
Learnings
Prediction accuracy varies wildly with market conditions, our signals work great when trends are clear but struggle in sideways action. Makes sense since we built them to catch momentum, not range-bound trading.
What's Next
- Finish the conditional logic for different market regimes in the prediction model. Want to ship a version that admits uncertainty instead of pretending to be accurate in all conditions.
- Continue partnership discussions and potentially start a technical integration prototype if things progress.
- Start thinking about how to present predictions to users.
Closing Thought
Building predictions is easier than building trust in predictions. The real work isn't in the algorithm, it's in showing users when to ignore it.
1
88
gorilli retweeted
🦍 Gorillionaire – Weekly Update
Week 3: January 12 – January 18, 2026
Summary
This week was all about getting deeper into the prediction engine. We kept grinding on the prototype that transforms our signals into actual price forecasts, the math is getting interesting but also messier than expected. Also started exploring a potential partnership that could change how we think about signal distribution.
Highlights
- Prediction Model Deep Dive: Continued building the prototype that uses our signals and tracked wallet data to predict token prices. Been testing different time horizons, 4-hour predictions show 68% accuracy, but 24-hour drops to 52%. The challenge isn't the signals themselves, it's handling the noise from low-volume Monad tokens where one $50k trade can move price 15%. Currently experimenting with different types of confidence scores.
- Data Pipeline Optimization: Had to rebuild how we feed historical data into the model. Original approach was pulling everything fresh each run, which took 12+ seconds. Switched to incremental updates with caching, now predictions generate in under 3 seconds. Small win but feels way better for users.
- Partnership Exploration: Started conversations with a potential partner who's interested in our signal infrastructure. Can't share details yet, but it's the kind of integration that could put our signals in front of 10x more users. Still early discussions, but the technical integration looks feasible.
In Progress
- Prediction Algorithm Refinement: The core model works, but needs smarter handling of market conditions. Working on conditional logic that adjusts the model based onthe overall Monad ecosystem volatility. Probably need another week of backtesting.
Learnings
Prediction accuracy varies wildly with market conditions, our signals work great when trends are clear but struggle in sideways action. Makes sense since we built them to catch momentum, not range-bound trading.
What's Next
- Finish the conditional logic for different market regimes in the prediction model. Want to ship a version that admits uncertainty instead of pretending to be accurate in all conditions.
- Continue partnership discussions and potentially start a technical integration prototype if things progress.
- Start thinking about how to present predictions to users.
Closing Thought
Building predictions is easier than building trust in predictions. The real work isn't in the algorithm, it's in showing users when to ignore it.
Built in assistant that helps you with any problem or question on the platform, signals made easy! 🏌️
Mainnet Chronicles 📜
🛠️Your favorite signal and analysis got better on mainnet, you don't have to be slow anymore.
- Indicator based signals and analysis
- Token dashboard and token analytics
- APIs
Mainnet started slow but your trades won't!
Ggorilli
Gmonad!
1
1
123
Mainnet Chronicles 📜
🛠️Your favorite signal and analysis got better on mainnet, you don't have to be slow anymore.
- Indicator based signals and analysis
- Token dashboard and token analytics
- APIs
Mainnet started slow but your trades won't!
Ggorilli
Gmonad!
gorilli retweeted
🦍 Gorillionaire – Weekly Update
Week 2: January 5 – January 11, 2026
Hey @DeltaV_xyz, here is the @gorillionaireAI weekly update.
Summary
Started the year by shipping actual features. We plugged Kommander AI (an agent product we built) into Gorillionaire so users can finally ask questions without leaving the app. Also realized our Discord was getting messy with multiple projects, so we rebranded it to Gorilli to cover everything we're building.
Highlights
- Kommander Integration: Got our AI chatbot working inside Gorillionaire. Users can now ask "wtf does this signal mean" and get actual answers. No more switching between Discord and the app.
- Discord Rebrand: Changed the server from Gorillionaire to Gorilli. We're building multiple products now, so it makes sense to have one place for everything instead of fragmenting the community.
- Community Call Setup: Scheduled a call with users next week. Need to hear directly what's working and what's not.
- We created the @gorillistudio X account to gather broader gorilli activities.
In Progress
- Price Prediction Models: Working on something interesting: using our signals to predict future token prices. Early tests look promising but need more data to validate.
- Community Call Prep: Putting together questions for next week's call. Want to make it useful, not just another "roadmap update" thing.
Learnings
- Having an AI answer questions in-app is way more useful than we thought. Users actually use it instead of figuring out the product by themselves.
- Rebranding to Gorilli makes way more sense, we're building multiple things and want to share everything with our community, not just Gorillionaire updates. One place to share all our experiments and ship logs.
What's Next
- Ship the price prediction feature once we're confident it's not just noise.
- Run the community call and actually listen to what people want
- Feed more Gorillionaire-specific data into Kommander so it can answer better questions
- Figure out how to share data between our different products without creating a mess
Closing Thought
Sometimes the best features are the obvious ones. Users had questions, now they have a chatbot.
Simple.
happy to have chosen @0xProject as a partner in this
1
20










