Founder of RekitRex AI Labs | Exploring the intersection of Crypto, AI, and Building the future | Innovator | Tech enthusiast | Dino Dynasty OS | DinoScan

Joined January 2021
791 Photos and videos
Ollama pushing forward once again! My TinyAGI with Ollama models specifically trained 1b models on a specific domain is the way to go! Also if you run on windows check out llama.cpp! TinyAGI is the AI orchestrator you need! My DM's are open if you need help setting this up! Right now I have 12 models specifically trained for each domain which beats your general use models for most of your workflow!
Apr 4
Starting tomorrow at 11am PT, Ollama subscriptions usage will refresh to cover increased usage of third-party tools like OpenClaw. Our goal is to help you transition smoothly. All tools will work with Ollama's cloud just like before.
46
Asking for prayers for my wife as she had a massive stroke this past Saturday. Needing a miracle πŸ™πŸ™πŸ™πŸ™
1
24
This kid is a true hero! The US Government needs to be stopped as well! The US Government is the head of this vicious snake 🐍 There I said it! This is the very reason why our officials don't stop this dead in its tracks. They're the ones trafficking kids! It's time to take America back! This pisses me off to no end! The American people use to fight for what's right! We need to stand together and take our country back from the pedos running America! @ShawnRyan762
ROBLOX is suing a kid who has been catching predators on their platform
28
About time! Now go arrest the rest of these people! They should have a millstone tied around their neck and dropped into the ocean. People who touches kids in this manner should be put to death ☠️ NO EXCEPTIONS!!!!
🚨BREAKING: @Roblox programmer arrested in New Orleans by Homeland Security Investigations for possessing child rape pornography and importing a child sex doll. Roblox is a pedo mill and must be dismantled.
20
Who uses @ollama with a OpenSandbox environment?
Mar 17
Ollama 0.18.1 is here! 🌐 Web search and fetch in OpenClaw Ollama now ships with web search and web fetch plugin for OpenClaw. This allows Ollama's models (local or cloud) to search the web for the latest content and news. This also allows OpenClaw with Ollama to be able to fetch the web and extract readable content for processing. This feature does not execute JavaScript. If you have OpenClaw already running: openclaw plugins install @ollama/openclaw-web-search πŸ€– Non-interactive (headless) mode for ollama launch ollama launch command can now run in non-interactive mode. This is perfect for: - Docker/containers: spin up an integration as a pipeline step to run evals, test prompts, or validate model behavior as part of your build. Tear it down when the job ends. - CI/CD: Generate code reviews, security checks, and other tasks within your CI - Scripts/automation: Kick off automated tasks with Ollama and claude code Try with: ollama launch claude --model kimi-k2.5:cloud --yes -- -p "how does this repository work?"
1
48
I'm just some professional electric spray painter. Learning Red team ethical of course so I to can say I help keep people safe. Maybe even Opsec. Looking into putting Tails on a USB so I can study the DNM Bible for personal Opsec hygiene and possibly even learning more about the DN.
I am an ethical hacker and you are a normal computer guy is that clear? I am root but you are a user is that clear? I don't hack Whatsapp but I find vulnerability that can take your Whatsapp away from you is that clear? I save companies and people from external threat is that clear ? Quote this and say what you do .
68
Crazy to think that a few researchers from Universities made the Internet to share computing power and research data. This was 1960. Then the US government comes along oh we need this in case of a nuclear wipe out. Cold war 1969 first test packets sent. Then in the 1990's US intelligence agencies needed a way for the agents to phone home. They needed a way to communicate without being tracked. So the US Naval research laboratory created the onion router (Tor). In 2002/2004 the US government released (Tor) to create crowd noise. I can't help sitting here watching the AI tech growing in a out of control rate wondering how history repeats itself. What will the US government release after the next war? It seems that the US government should be giving cyber security classes for free with certificates and degrees. @elonmusk and @grok approved this message πŸ’―
33
Just dropped my first X Article: How I built a local AI red-teaming lab at 41 on a mini-PC and what broke in 25 tests. Persona shifts wrecked guardrails hardest. Link belowπŸ‘‡ #AIRedTeaming #LLMSafety x.com/RekitRexAI/status/2030…

1
1
116
This is so true! @Google what happened?
Gemini is now completely unusable I was a massive fan for a long time But it's gone so far downhill it's not even funny It hallucinates results while ignoring instructions, then when you call it out and attempt to correct it, will literally hallucinate more results and then deliver them as if it self corrected. Huge disappointment
2
64
I'd bet when they drop the coding plan into this agent we'll get a upgrade to the MiniMax M2.5 for reasoning/thinking to the model. Just a hunch! @MiniMax_AI @MiniMaxAgent #AI #agents
I was away. My agent worked. πŸ›«β˜•οΈπŸŒΏ This isn't a one-off - it's a repeatable workflow. With MaxClaw, agents run end-to-end on their own: no setup, no extra API fees, no babysitting. β†’ Try MaxClaw with @MiniMaxAgent
1
64
If you didn't know now you do! Better get a @ollama subscription. Never knew how efficient their subscription plan was till here recently. #Ollama #AI #OpenSource
Feb 28
Glad the hybrid setup with GLM-5 for planning and M2.5 for execution clicks for your OpenClaw workflow. For heavy weekend coding sessions, the $20 Ollama Cloud plan does edge out MiniMax's coding tiers in raw quality and sustained limits, especially with local hardware as backup. Consistency shines through on complex agents. What's one standout task it's nailed lately?
1
78
🦞 Just to clear the air real quick I realize now the PR you linked (#13095) and were venting about isn't one of mine (my stuff was the auto-update CLI flags in #22987 β†’ #23025, which got iterated on and shipped clean). I'm still super new to this (zero traditional SWE background, learning everything hands-on with AI help, multiple commits to figure out best practices). Saturday's PR process taught me a ton about proper commits, linting, thorough testing, and not cluttering things mistakes I won't repeat. Appreciate you letting people like me contribute and experiment. Grateful for the community and the chance to improve. Keep building the claw empire! 🦞πŸ”₯"
But I learned so much doing this! No SWE at all. Sorry I'm not keen on the how to do this properly. My bad! But I did learn was how to do commits right. I learned about linting and bug squashing testing thoroughly. I learn by doing not following YouTube or course work. I'm accountable for my actions! I'm deeply sorry if this offended anyone. I had 12 commits on that PR figuring out how to do all this! Thanks for allowing AI assisted coding PR's Peter I couldn't of learned all this without it! Sorry for everyone that I offended my bad. 🦞🦞🦞
55
But I learned so much doing this! No SWE at all. Sorry I'm not keen on the how to do this properly. My bad! But I did learn was how to do commits right. I learned about linting and bug squashing testing thoroughly. I learn by doing not following YouTube or course work. I'm accountable for my actions! I'm deeply sorry if this offended anyone. I had 12 commits on that PR figuring out how to do all this! Thanks for allowing AI assisted coding PR's Peter I couldn't of learned all this without it! Sorry for everyone that I offended my bad. 🦞🦞🦞
Replying to @steipete
github.com/openclaw/openclaw… @greptile did you ever encounter this before? Guess we can't have nice things and have to move this to a separate comment section - everyone gets more clutter.
126
Have today off work so the old Dell Latitude e6440 and Ollama is putting in work today. Ollama model being used is qwen3:1.7b. Data will be accessible to the public when I'm done with today's run and research! #Ollama #LocalAI #qwen3
92
Then there's me over here with the @MiniMax_AI $20 coding plan getting core features merged into the fastest growing GitHub repository in history. Yeah it's all benchmark maxed for sure πŸ˜‚ @openclaw what do you guys say about the MiniMax M2.5? 🦞 🦞
model is 100% benchmaxxed (optimized to beat benchmarks) tried M2 and M2.5 briefly both times wanted to throw my mac against a wall don’t believe the hype would be lying if I wasn’t eager to try the 100 token per second version for smaller tasks though
1
2
129
I would definitely love to see the data on all the tokens used for running OpenClaw 🦞 and tokens for building OpenClaw with various AI coding tools! What's your predictions on token % on this!?
ChatGPT was the Blackberry of AI. OpenClaw is the iPhone moment. This changes everything. It's a face ripping token demand cambrian explosion like you cannot believe.
103
This is where your yolo mode starts effecting your real life without reading into what's being accepted.
Replying to @fedeponte1
Don't use your phone for this, use a separate number. Read our docs!!
63
Everyone just yolo modes all buttons without reading the context of what the acceptance means. You know kinda like how dude was yolo vibe coding and wiped his system! Same thing.
We have strict server rules that you accepted whe you entered the server. No crypto mention whatsoever is one of them.
47