0xBralca 📈 📣 | BMM

966 posts

0xBralca 📈 📣 | BMM banner
0xBralca 📈 📣 | BMM

0xBralca 📈 📣 | BMM

@0xBralca

💎 Find hidden gems before they PUMP ➡️ https://t.co/INfHiOnfEO 📈 Join the waitlist for the Bull Market Multiplier indicator ➡️ https://t.co/6dGh67CMdh

on-chain Katılım Ekim 2020
489 Takip Edilen3.1K Takipçiler
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
It's so cool that now in @afkode_ai I can select different models based on the type of task and complexity. I did 2X more medium complexity backend task than high complexity but I spent roughly the same amount of tokens. Also the medium complexity tasks completed on avg 2X as fast as the high complexity because I am using a cheaper faster model for those. Before this I would have just used the best possible model for all tasks and would have ended up spending A LOT more tokens and A LOT more time!
Alex I/O tweet media
English
1
1
4
187
0xBralca 📈 📣 | BMM retweetledi
Afkode
Afkode@afkode_ai·
🔊 SOUND ON > All you see and hear is made only with AI
English
2
1
2
8.5K
Alex I/O
Alex I/O@alessiocarra_·
You can now assign 21 different models for the right task in @afkode_ai All of them will have the correct context to complete their job automatically. During autonomous execution you can have - Low complexity Frontend Tasks -> Kimi 2.6 - Low complexity Backend Tasks -> Deepseek V4 - Medium complexity FE -> Sonnet 4.6 High thinking - Medium complexity BE -> GPT 5.5 medium thinking ... You get the point. The best way to save on your AI bill and get the best model do the job you need
English
2
0
3
1K
0xBralca 📈 📣 | BMM
0xBralca 📈 📣 | BMM@0xBralca·
This is amazing tbh
Alex I/O@alessiocarra_

You can now assign 21 different models for the right task in @afkode_ai All of them will have the correct context to complete their job automatically. During autonomous execution you can have - Low complexity Frontend Tasks -> Kimi 2.6 - Low complexity Backend Tasks -> Deepseek V4 - Medium complexity FE -> Sonnet 4.6 High thinking - Medium complexity BE -> GPT 5.5 medium thinking ... You get the point. The best way to save on your AI bill and get the best model do the job you need

English
0
0
1
59
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
Imagine being stuck for 270 minutes on testing instead of letting @afkode_ai do all the work and wait for the full test suite to run and then check the logs and then fix the issues and then keep doing that until all tests pass... Just imagine...
English
0
1
4
491
0xBralca 📈 📣 | BMM retweetledi
Afkode
Afkode@afkode_ai·
Meet Afkode: the builder-first harness for autonomous feature delivery. Give it a complex feature. It plans, builds, tests, reviews, and keeps going until it works. Run parallel projects, mix models, validate with real tests, and ship more with less agent babysitting.
English
0
3
6
52.5K
0xBralca 📈 📣 | BMM retweetledi
Afkode
Afkode@afkode_ai·
🚨 Opus 4.7 is now available in Afkode. You can assign it to 16 different activities across planning, implementation, review alongside any other model or provider of your choice to work on the same feature at the same time. Check it out: afkode.ai
Afkode tweet media
English
0
2
3
160
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
Spent 15 minutes to scope a HIGH complexity feature. The app will plan this for the next 2 hours. Then proceed to implement it in the next 4 to 6 hours. Fully autonomously. To give you an idea of the complexity I am throwing at it, the description is in the first comment
Alex I/O tweet media
English
1
1
2
161
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
Screw it! I am building a new IDE. I know most people are already rolling their eyes BUT After using Cursor, Claude Code, Antigravity, Codex, etc. I felt they weren't serving me they way I wanted. They are not meant for founders, they are meant for devs. I want to be able to plan the work (where I spend most of my time) and then while the plan is being built, go and do something else. But baby sitting the models makes that impossible. The problem is that leaving Claude or Cursor in YOLO mode doesn't really work cause they will mess up the context in ways that I have no visibility or control over. This new IDE will solve all of that, with novel solutions to context management that none of the big platforms have. It will be possible to let the models run for hours without supervision and loss of quality due to Context Compacting or Context Rot or poor Context in general. Stay tuned!
Alex I/O tweet media
English
13
2
32
2.5K
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
BAMBAMBAM! Finally starting to show up in top placements on Google! Some time ago I set the goal to rank in top positions for "Context Engineering". It takes a lot of work and patience to build the basics for this. weeks and weeks where you keep pushing and nothing happens. BUT It feels so good when it finally starts paying off!
Alex I/O tweet media
English
1
1
3
176
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
Many people have asked me to make a video and explain how the Context Engineer works So this is it! What it's covered: • How to install and set up the Context Engineer MCP • How to start and run a planning session • How to generate PRDs and Tech Specs directly from your ideas and code • How to feed the right amount of context at the right time • Best Practices to follow when implementing the plan If you have any feedback just comment below!
English
1
1
4
948
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
> "I want AI that don't always agree with me" I AM COOKING!!!
Alex I/O tweet media
English
0
1
3
157
0xBralca 📈 📣 | BMM retweetledi
Alex I/O
Alex I/O@alessiocarra_·
SEO starting to deliver something! Some good keywords we rank high Impressions spiked recently Avg. Position needs to be <10 What else?
Alex I/O tweet media
English
1
1
6
208