Prashant Raval

I’m Prashant, a technology professional with 15+ years of experience in the IT industry, specializing in website and Android app development.

[App Hunter  @Sn0wbrave]

إذا كنت تبحث عن أداة لالتقاط الشاشة على ويندوز، فهذا المشروع يستحق التجربة.

WinShot أداة مجانية ومفتوحة المصدر تتيح التقاط الشاشة، إضافة الشروحات، القص، ثم التصدير بواجهة حديثة وسريعة. تعمل على Windows 10 و11 وحجمها صغير.

[Geek Lite @QingQ77]

面向工业物联网（IIoT）的 MQTT 诊断工具，能连任意 broker、实时浏览主题树、查看负载、绘制 JSON 指标图表，并原生解码 Sparkplug B、模拟 EoN 边缘节点，全程开源、不依赖云端。

https://t.co/qayUnYX9HE

[LTX.io @ltx_io]

Most reframe tools crop, downsize, or hallucinate.

Introducing the LTX Reframe API.

Expand any video to a new aspect ratio. The original frame stays untouched. The model generates only what’s beyond the edges.

• Full source length, up to 60s
• Native 1080p
• 16:9, 9:16, 1:1, 4:5
• Built on LTX-2.3

One master reframed for every placement.

[⚡AI Search⚡ @aisearchio]

LingBot-World 2.0 by @robbyant_brain is the most impressive open-source world model I’ve seen this year. It generates interactive worlds which you can explore and take action in real time.

> Add actions, events, weather changes, object interactions, and more
> 720p @ 60 FPS
> A 60-minute uninterrupted session across 20 scenarios with no perceptible quality decay
> Agentic "Brain + Cerebellum" system that continuously proposes new events as you interact

Open source, 14B model available. Lightweight 1.3B variant for consumer-GPUs coming soon

Github: https://t.co/FfpBXZTkpM
Try it online: https://t.co/Dmj9UXq1qQ
Weights: https://t.co/qPUP7rTqKh
Website: https://t.co/B3N1xZmeJX

[Tengfei Wang @DylanTFWang]

⚡️ Tencent HY-World 2.1 is HERE!

Not a video. A world. 🪐

3 months just after 2.0, we've supercharged everything:
✨ Cleaner geometry
✨ Sharper rendering
✨ Larger explorable range

It's a REAL 3D world you can Walk through and Interact with.

[Hugging Models @HuggingModels]

Meet Dolphin3-Cyber-8B-GGUF, an 8-billion parameter LLM fine-tuned for cybersecurity. It excels at penetration testing, security analysis, and threat detection. A must-have for ethical hackers and red teams.

[Jun Song @jun_song]

SuperGLM-5.2-abliterated is finally available 🚀

The strongest local AI that you can run on your hardware.

- uncensored from Nvidia NVFP4
- Fixed broken points from abliteration
- enhanced overall performance

Only available with nvfp4 now ⬇️

[Akshay 🚀 @akshay_pachaar]

the four types of agent loops.

loop engineering keeps getting talked about as one thing. it's actually a choice between four structures, and each one fits a different kind of task.

it means designing the system that steers the agent, instead of steering it yourself move by move.

that system always answers two questions. what starts a run, and what decides the work is done.

in a hand-run session you answer both yourself, every single time. each loop type moves more of that into the system.

here's each type, what triggers it, and when to reach for it.

1) turn-based.

triggered by a user prompt. the agent gathers context, acts, and checks its work inside a single turn, then a human reviews the output and writes the next prompt.

use this when requirements are still forming and every output changes what you'd ask for next.

2) goal-based.

triggered by a /goal command carrying success criteria and a budget, like "get the homepage Lighthouse score to 90, stop after 5 tries." when the agent tries to stop, an evaluator model checks whether the goal is met, and a no sends it back to work.

use this when the outcome is measurable but the path there isn't worth your attention.

3) time-based.

triggered by a clock. an interval fires, the agent runs a fixed prompt like "check the PR, fix CI," then waits for the next tick. /loop runs on your machine, /schedule moves it to the cloud so it survives a closed laptop.

use this for recurring work where the task is known in advance and only the timing repeats.

4) proactive.

triggered by an event or schedule with no human present. a routine watches a channel, and when something needs handling it spawns a workflow with a triage agent, a fix agent, and a reviewer that adversarially judges the work before the task closes.

use this for standing responsibilities where you can't predict what will come in, only that something will.

each type hands off one more job than the last. turn-based keeps both with the human, goal-based automates the checking, time-based automates the trigger, and proactive automates both while deciding the workflow shape at runtime.

so the mapping question isn't which loop is most advanced. it's whether your task is exploratory, measurable, recurring, or standing.

the more you hand off, the less you babysit.

I wrote the full breakdown on loop engineering. the article is quoted below.

[ハカセ アイ(Ai-Hakase)🐾最新トレンドＡＩのためのＸ 🐾 @ai_hakase_]

【Gemma 4 × React Native！完全オフラインで動く超高速オンデバイスAIの実装】
👉 https://t.co/xJEYFDu1So
モバイルAIの未来がやってきました！Googleの最新LLM「Gemma 4」を、スマートフォン上で100%完全にオフライン動作させる画期的な技術が注目を集めています📱✨

なんと、クラウドAPIへの依存をゼロにすることで、運用コストは驚きの「0ドル」！データが一切端末の外に送信されないので、セキュリティやプライバシーを最優先にしたい医療や金融などのアプリ開発にぴったりですね。

この仕組みを支えるのが、Metaの推論エンジン「executorch（エグゼキュータッチ）」をReact Nativeに統合した「react-native-executorch」というライブラリです。モバイルデバイス固有のパワーを引き出すため、Androidでは「Vulkan」、iPhoneでは「MLX」のバックエンドを利用して、GPUやNPUをフルに活用したハードウェア加速を実現しています。

さらに、単なるテキストチャットだけでなく、ローカル環境での画像認識（Vision）や自動で機能を呼び出すツール実行（Tool Calling）にまで対応しているのが凄すぎます！
開発向けには「useLLM」という使いやすいReactフックも用意されており、アプリのコンポーネント内で直感的にAIの動作を制御できちゃいます。性能はスマホ本体のスペックに依存しますが、通信環境を気にせずサクサク動くプライベートなAI機能を作りたい時には、間違いなく最強の選択肢になりそうですね！💡

#ReactNative #Gemma4

[apolinario (poli) @multimodalart]

GPUs for all 🤗

Creating ZeroGPU demos or apps is now available to ALL @huggingface users

tell your agent: "Build a HF ZeroGPU demo for this model"

New Space: https://t.co/kSdMdiemPP
Agent skill: https://t.co/JuQVcthoM0

[Leonard Rodman @RodmanAi]

10 AI Repositories That Just Went Viral on GitHub:

Bookmark these before everyone else does.

1. Meetily
An AI meeting assistant that runs entirely on your device. Powered by Parakeet/Whisper for fast transcription with zero cloud uploads—perfect for private meetings.
https://t.co/mG2T7AHhHI

2. agent-skills
A library of battle-tested skills for AI coding agents like Claude Code. Adds real-world workflows for testing, debugging, reviewing, and refactoring.
https://t.co/5t5lLf3UK3

3. system_prompts_leaks
A massive collection of leaked system prompts from ChatGPT, Claude, Gemini, Grok, and more. Great for understanding how leading AI models are actually instructed.
https://t.co/GGldCif4pl

4. OfficeCLI
Give AI agents direct access to Word, Excel, and PowerPoint through a simple CLI. No Microsoft Office installation required.
https://t.co/ZNyvK3nLSe

5. Orca
Manage multiple AI coding agents from one interface. Launch parallel tasks, track progress, and work across desktop and mobile.
https://t.co/UrP8x1Iytd

6. OmniRoute
One API endpoint for 230+ AI providers. Includes access to free model tiers and smart token optimization to reduce costs.
https://t.co/P2R2KRmY9e

7. Claude Video Reader
A tool that lets Claude understand videos by automating downloads, extracting frames, generating transcripts, and packaging everything into AI-friendly context.
https://t.co/r0a5kzUJey

8. page-agent
Control web pages using natural language. AI agents can click buttons, fill forms, and navigate websites without custom automation scripts.
https://t.co/OHr1v2nCHy

9. Astryx
Meta's open-source design system built for both humans and AI agents, making UI development more consistent and AI-native.
https://t.co/1AaE6eqlaW

10. CubeSandbox
A fast, secure sandbox for AI agents to execute and test code in isolated environments with near-instant startup.

Which repo are you starring first?

[Denzii 🕊️ @denziideng]

25k Star，港大开源一款 AI 个性化辅导私教神器，在 GitHub 上杀疯了！🔥

今天带来DeepTutor，港大开源的AI个性化辅导私教！GitHub 25k🌟，100天内爆火了。它把出题、解题、研究、测验、知识管理全整合成一个智能体工作区，让AI真正陪你终身学习。

📌 六大模式无缝切换：Chat、Quiz、Research、Solve、Visualize、Mastery Path共享一个Agent引擎

📌 强大记忆系统：三层记忆+Memory Graph可追溯，上下文始终跟随你

📌 Partners与My Agents：轻松接入本地模型，独立Persona和技能，记忆隔离

📌 知识中心+活书编译：多引擎RAG支持PDF等文档，笔记一键生成可对话书籍

📌 Co-Writer协同写作：智能编辑、标注、TTS，一键导出Markdown

👋使用方法简单直接：pip安装或Docker一键部署，浏览器打开本地界面就能开始聊天学习。支持本地Ollama模型，配置好后直接用自然语言下指令，AI自动规划并执行。

🔗GitHub地址：https://t.co/pMVSMiMEER

强烈推荐给学生、终身学习者和想做AI教育工具的开发者，有个真正懂你、记得住的AI私教，还报啥辅导班啊！😄

#开源工具 #AI教育 #个性化学习 #DeepTutor

[Encore @encoredotdev]

Introducing the Database Explorer - now in the cloud.

Browse your schema and run queries against your deployed databases right from the dashboard, with read or write access scoped to each environment.

Give it a go → https://t.co/nIexCTDDJP

[GitHub Projects Community @GithubProjects]

LangChain-Chatchat lets you run a full RAG pipeline privately using open-source LLMs and local knowledge bases.

- Supports ChatGLM, Qwen2, and Llama3 via Xinference or Ollama
- Fully offline deployment with no external API dependencies
- Built on LangChain, FastAPI, and Streamlit for modular use
- Works with major open-source embedding models and vector databases

Explore it here:
https://t.co/u0PNf3SF6C

[Zane Chen @chenzeling4]

Joining a new team and staring at 200K lines of code you didn't write.

Understand-Anything turns any codebase into an interactive knowledge graph. Search it. Click through dependencies. Ask questions in plain English. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI.

This is what onboarding should feel like.

⭐ 73.7K #DevTools #OpenSource

https://t.co/hLTAtvAIU9

Follow for daily dev finds 🔔

[Simplifying AI @simplifyinAI]

This free tool is the open-source alternative to Google Photos.

It's called Immich.

Your photos and videos live on your own server. Auto-backup from your phone, facial recognition, shared albums, and memories.

Everything Google Photos does, without the subscription.

iOS and Android apps included.

[Blaine Brown @blizaine]

Maestro v1.1.3 is out!

Maestro is a free, open-source creative studio that lets you generate videos, images, music, voices, and sound effects on your own computer without having to piece together a collection of complicated AI tools, using the latest open SOTA models.

Director mode can take (or generate) a song or story idea, plan the scenes, generate the visuals and video clips, and assemble them into a complete music video or short film.

New features
• Linked Model Folders: point Maestro at your other installs (like Wan2GP) and reuse their checkpoints and LoRAs instead of re-downloading hundreds of GB. One-click scanning of your Pinokio apps, strictly read-only, and AI LoRA guides work on linked LoRAs too.
• Krea 2 image models (Raw and Turbo).

Quality and performance
• Major quality fix for LTX-2 Dev and 10Eros: a leaked sampler setting was making output blurry and oversaturated. This was the root cause of the "Dev models look bad" reports.
• The STG slider actually works now (it was silently doing nothing).
• Performance Auto-Tune gives audio its own memory profile: cards with 12 GB+ VRAM now unlock the fast ACE-Step song decoder instead of a slow fallback. Re-run Auto-Tune after updating to pick this up.
• ACE-Step song generation no longer looks like it hangs forever. The runaway progress counter was a display bug; it now reads honestly.

Director mode
• Regenerating a clip from the Pipeline Dashboard keeps the original song, conditioned on the exact segment of the soundtrack that clip covers.
• Re-join works end to end: full-length clip reruns, original soundtrack overlaid, nothing drifts out of sync.
• The gallery refreshes automatically when new clips or rejoined videos are saved.
• Fixed bogus "Generate N missing" counts, broken start-image thumbnails, and LoRA recommendation colors showing wrong on some themes.

Click the Update button in the Pinokio launcher.

[Excalidraw @excalidraw]

Our Excalidraw+ MCP is getting stronger by the day!

Diagrams and arrow connections now improved.

[Miguel Ángel Durán @midudev]

Un agente de código con IA totalmente gratis.
Sin suscripciones, sin pagos ni tarjeta de crédito.

Tiene disponible DeepSeek, Kimi K2.7 y MiniMax M3.

¿Y el truco? Muestra pequeños anuncios de texto entre las respuestas del agente

https://t.co/KKD7wwjwzi

[Manish Kumar @manixh]

Linux is PEAK ‼️