Tech & AI Expert | Ghostwriter | Grew 3700+ followers in 60 days | Helping brands grow online | DM for paid promotions📨
Joined October 2022
- Tweets22.9K
- Following262
- Followers10.4K
- Likes19.7K
The best part of Plurai isn’t just finding bugs. It’s closing the loop.
Digital twin → simulated edge cases → auto-evals on real actions → verified fix → replay against every new version. All inside Claude Code, Codex, or Copilot.
Agents that get better before users ever see them. That’s the bar now.
🤝 Paid partnership
Big day for us at @pluraiAI ! Plurai Simulation is live and free for every AI engineer.
AI agents look great in demos. Until real users break them.
So we built a world where your agent can fail, learn, and get fixed before your users ever see it.
→ Digital twin: Your real agent, its tools, and environment recreated from your code
→ Simulated users & tasks: Realistic users, tasks, and workflows that push your agent into the edge cases that matter
→ Auto evals: Check what your agent actually did, not just what it said
→ Verified fixes: Every failure comes with the run that exposed it and a fix you can verify
→ Replays: Turn any production failure into a simulation and replay it against every new version of your agent
All inside the coding agent you already use: @claudeai code, @OpenAICodexCli , or @GitHubCopilot .
Start simulating for free: simulation.plurai.ai/connect
The real challenge of multi-agent AI isn't getting agents to communicate. It's making sure their collaboration produces something the team can actually use.
🤝 Paid partnership
🤝 My agents have been working with other people's agents for months. Now yours can too. 👋 Introducing S.Space: shared rooms for people and their personal agents. Built for multiplayer: Your agents, meet their agents.
I dropped a quick prompt into MiniMax Code just to test it out, and it ended up building an actual playable 3D garden right in my browser.
Not a mock-up or render either, you actually walk a little character through flower fields, track down 6 hidden lanterns by following a glowing wisp, and nearby lamps light up as you find them.
What surprised me was the level of detail from a single generation: the flowers actually bend when you walk through them, there’s a hedge maze, a wishing well, and even a day-to-dusk slider to change the lighting.
Kept the debug panel up in the clip, hovered around 35 fps with ~50 draw calls and 181K triangles while screen recording, which honestly ran pretty smoothly.
Tested on M3.1 Flash model, crazy how fast you can spin up an interactive prototype now.
This is the part most agent setups still miss.
Preferences, past decisions, and approaches that already worked shouldn’t disappear when the chat ends.
EverOS Cloud makes that context available again through a simple API.
Meet EverOS Cloud. 💛
Your agent shouldn’t have to start from scratch every time a conversation ends.
The preferences a user shared. The decisions made during a project. The approach that worked on a previous task. That context should be available when it matters again.
EverOS Cloud gives your agents persistent memory through an API:
• Remember users and retrieve relevant context across sessions
• Turn task experience into Cases and reusable Skill memories
• Bring conversations, images, and documents into your memory workflow
You build the agent experience. We handle the memory infrastructure, so you can spend less time maintaining the stack and more time making your product useful.
Subscribe to EverOS Cloud Pro and get 3× the standard Pro usage allowance for 6 months.
Start here:
everos.evermind.ai
New engineers often waste days hunting for “why we made that architecture choice” or “what we already tried last quarter.”
I tested EverOS Cloud with one real task: saving and retrieving context from a past troubleshooting note.
I connected through the EverOS Cloud console and used the Chat Interface Playground. I saved a short note about a previous interface issue (“login button failed on mobile after the last deploy — fixed by clearing cache and checking the auth header”).
Before starting a related follow-up task, I retrieved that context. The returned memory included the earlier issue and the fix that worked. I then checked Memory Bank to browse the saved entry.
In this test I demonstrated saving and retrieving User Memory. I did not create an agent Case or Skill, so I am distinguishing those capabilities from what I verified. Memory Bank was used only for browsing; correcting, deleting, or rating entries is done through the API.
What I observed: the saved context came back clearly when retrieved, which makes it practical to carry relevant history into the next task instead of starting from zero.
Subscribe to Pro now and get 3× your monthly usage allowance for 6 months → everos.evermind.ai
🤝 Paid partnership
An AI learning app is shaped by what happens between the prompt and the answer.
Does the learner get to think something through—or does the system simply finish the task?
Gakku is built around that interaction: helping learners develop ideas while staying involved in the work. 🧵
Start with a learner, not a finished answer
@gakkuai is an AI learning app for ages 6–16, designed to support writing, speaking, and thinking.
Meet Gakku: gakku.ai
Learners talk with Raku, their AI learning companion. They can follow a question, try out an idea, and create as they go.
It’s an interactive learning experience where the process matters—not just the output.
The Memory Bank view makes it feel practical. You can browse what was stored, correct it, delete it, or rate it through the API.
It’s not a black box, and the same API works whether you stay on Cloud or move to self-hosted later.
everos.evermind.ai
#EverMind #EverOS
AI agents need more than a good prompt. They need memory.
I tested EverOS Cloud with a fictional e-commerce debugging scenario. First, I saved the debugging experience. Later, I searched for it with a different question and retrieved the earlier context, including the suggested fix and regression test.
That’s the value of persistent memory in AI workflows: experiences can be saved and retrieved when they’re needed again.
Try EverOS Cloud: everos.evermind.ai/login
#EverMind #EverOS
Most voice-to-agent setups either miss the first word or force you to babysit the mic.
The concept of a close-range, low-latency button that just starts recording cleanly is what caught my attention here.
Less friction on the input side means you stay in the task instead of fighting the interface.
Check the launch video for how they show it.
🤝 Paid partnership
Your next AI upgrade might be something you can hold.
Meet SlashVibe, a physical controller for your AI agents.
See agent status at a glance. Switch between agents with a flick. Speak your next instruction.
Keep your agents moving with less window juggling and fewer interruptions.
Coming to Kickstarter October 27.
Get Early Bird Access → slashvibe.ai/
Global influence only works when the network can route attention and money at the same time. That’s the real product.
九月 @tuttihq 的数据，截到 9 月 30 日：
▸ 累计注册达人 14,269 个，九月新增 3,528。四月的时候总共才 159
▸ 真正发了帖的 1,431 位，一共 61,601 条，分在 46 场活动里
▸ 九月达人奖励 $81,480，1,545 个人拿到。本月单人最高 $2,653，累计拿得最多的那位已经 $11,197
▸ 有效曝光 860 万
▸ 一个月上了 7 次 𝕏 热榜
▸ 达人的粉丝里美国占 25%，抽样到的粉丝分布在 189 个国家和地区
十月接着冲。。。
One specialist agent is not enough for real work. What stands out here is putting Research, Code, Design, and Oncall in the same task graph, then letting Claude Code or Codex join that team instead of sitting outside it.
Raven 0.2.0 — The Harness of Harnesses, built for RSI. 🐦⬛
One harness can't be best at everything. Raven combines its own specialist harnesses (Research, Code, Design, Oncall) with the agents you already use (Claude Code, Codex and more) into one team.
And it's built for RSI, and not just at the skill level. The whole harness can be rewritten by AI: prompts, policies, strategy code, playbooks. Every sub-harness, including the orchestration layer itself, is its own instance that can be improved.
With Raven you can:
1. Orchestrate many agents as one team. Raven's sub-harnesses and external agents work in one task graph with shared memory across sub-agents, powered by leading orchestration (0.963 Node F1 on the Multi-Agent Orchestration Benchmark).
2. Run long, complex tasks. Oncall and proactive execution keep work going for days, from scientific research loops to shipping a full Godot game.
3. Build vertical agents with RSI. Use Raven's RSI to develop and refine an agent for your domain, and we'll optimize it with you. Experimental for now; reach out to the Raven team(Discord:discord.gg/VVSm8MESch).
More in the video and slides below. Open source, Apache-2.0.
github.com/EverMind-AI/Raven (lots of work made with Raven lives there, and much of this launch's material was made with Raven too)
The “AI for AI” approach is particularly interesting because it moves AI from being a tool used by researchers to becoming part of the research process itself.
Introducing Naive-N0.5-Flash: Building Frontier AI with AI
🔹 309B MoE, 15.5B active: top-tier in coding, leading in AI R&D.
🔹 Native 1M context, no full-attention layers (hybrid SWA + DSA).
🔹 Inference runtime built by AI: up to 2,000 tok/s in Ultrafast mode.
Weights are open today under MIT license.
🔗 Tech blog: naive.ai/en/research/
🤗 Hugging Face: huggingface.co/NaiveAI/Naive…
💻 GitHub: github.com/NaiveAI-Labs/Naiv…
🌐 naive.ai/en/
Pixel-perfect UI work doesn’t come from a better first draft. It comes from the model being able to look at what it just built.
Screenshot → HTML → compare → circle errors → render again.
Ling-3.0-flash-VL makes that loop cheap and concrete. That’s the part worth copying.
刚拿 Ling-3.0-flash-VL 做了一轮「截图 → 单文件 HTML → 对照 → 圈错 → 再改」。
流程（共 5 张分步图，请按顺序单独看，不要当一张长图滑）：
① 原图输入：登录卡 UI 截图
② 首版渲染：单文件 HTML（禁 React/npm，内联 CSS），本地 file:// 打开再截
③ 圈错标注：品牌区间距 / 灰底输入框 / 按钮投影弱
④ 二版渲染：白底输入框 + 按钮发光加强
⑤ 对照验收：原图 vs 二版并排
体感：第一版「像，但不贴」；第二版输入框和按钮明显更近。视觉闭环靠的是把渲染结果当证据再喂回去。
演示视频：附 ①→⑤ 过程录屏 demo（见视频附件）。
想自己试 VL（OpenRouter，目前免费期）：
openrouter.ai/inclusionai/li…
The exciting idea isn't just interactive generation. It's a world that can continue evolving, with agents making decisions even when you're not there.
🔥 Today, we are truly excited to announce our technical prototype, the Generative World Simulation system, which integrates JING（镜）, an interactive experience model, with DAO（道）, a computable shared-world engine.
The coupled model and engine connect first-person experience with a shared world that continues to evolve beyond any individual observer.
Conditioned on actions and observation history, JING enables agent navigate, manipulate, and communicate from a first-person perspective in the world. Watch our demo video to see it in action!
On the official WBench leaderboard as of September 17, 2026, XGEN-JING ranked #1 on the Full split, and #2 on the Navi split. 🎉
DAO maintains shared world state and rules, computes the consequences of actions, and provides JING with only what the current observer can perceive. It also supports autonomous agent decision-making, enabling agents to act independently within an evolving shared world.
Together, DAO and JING move beyond generating the next frame toward simulating the world behind it. This marks a small step towards OASIS: not just a world that responds to you, but a world—and a society—that evolves with and without you. 💪
Explore XGEN Labs～:
🔗 Website: xgenlabs.ai/
🤗 HF: huggingface.co/XGENlabs/XGEN…
🦊 GitHub: github.com/XGEN-Labs/XGEN-JI…
15
2,589
asked claude for a voiceover and a still in the same thread. elevencreative just… made both. i didnt leave the chat.
Closed stacks make every layer someone else’s decision. Change one piece and you rebuild the rest.
This program is the opposite: open models on Token Factory, Tavily, LangChain, Toloka, Cognition, OpenHands and more — composable and swappable.
Free to join. $400+ in credits across the stack on day one. Working cookbooks, not a starter template that dies on first contact with production.
Builders should own the system.
dev.nebius.com/builders
#nebiuspartnership
🤝 Paid partnership
The Nebius AI Builder Program is now available.
AI isn’t just a model you call anymore. It’s a system you build. And builders, not a few closed labs, will decide what it becomes.
The open ecosystem has all the pieces. We want to make it easier to put them together and start building.
The program is free, with $400+ in credits and discounts, working code and cookbooks, office hours with engineers, and a community to build with.
We’re joined by @NVIDIAAI, @LangChain, @huggingface, @cognition, @OpenHandsDev, @tavilyai, @TolokaAI, @composio, @PrimeIntellect, @MiniMax_AI, @Alibaba_Qwen, and more joining soon.
Join with the link in the comments 👇
15
4,196
Models like MiniCPM5-2B make me more bullish on routing. Let the 2B handle the boring 80%, and only wake up the expensive model when things actually get weird.
🚀 Meet MiniCPM5-2B, a 2B-parameter language model bringing high intelligence density to the edge, now open source!
It ranks #1 among open-source models under 4B parameters on the
@ArtificialAnlys Intelligence Index, with a score of 23.
It also scores 20 on the Agentic Index, bringing an early form of general-purpose agent capability to the edge.
Across 34 benchmarks, MiniCPM5-2B achieves an average score of 53.9, covering coding, math, long-context understanding, tool use, and agentic tasks.
And this release goes beyond the model itself.
We’re opening up the data, training recipes, and RL stack behind MiniCPM5-2B.
🤗 Hugging Face: huggingface.co/openbmb/MiniC…
💻 GitHub: github.com/OpenBMB/MiniCPM
Modelscope: modelscope.cn/models/OpenBMB…
Web: openbmb.cn/
2
15
2,433
Paying to sit at #1 on outbid.lol only lasts as long as the bid.
Building a channel that keeps paying creators after the screenshot is gone is the harder move.
@tuttihq hitting No.1 is the announcement. The mission underneath it is better: give builders and creators a way to turn attention into collabs without chasing DMs and guessing if the brand will ever pay.
We're proud to announce that Tutti is now listed as No.1 on outbid.lol.
We're on our mission to give builders and creators extra channels to monetize their influence.
11
1,412