Fil public

Orchard is an open-source framework for the research community to train and evaluate AI agents across task types. It reduces complexity while supporting strong performance from smaller models by enabling researchers to reuse the same infrastructure:

via Microsoft Research

Tmux : le guide du multiplexeur de terminal

Découvrez tmux, le multiplexeur de terminal indispensable pour garder ses sessions, organiser ses fenêtres et ne plus jamais perdre son travail.

via slash-root.fr

FeyNoBg: A SOTA Model For Background Removal — Feyn

FeyNoBg beats or matches the current leaders across eight benchmarks. Alongside the model, we

via Feyn

OWASP Agentic Skills Top 10 | OWASP Foundation

Security Risks and Mitigations for AI Agent Skills

GitHub - YosysHQ/oss-cad-suite-build: Multi-platform nightly builds of open source digital design and verification tools

Multi-platform nightly builds of open source digital design and verification tools - YosysHQ/oss-cad-suite-build

via GitHub

GitHub - ghantoos/lshell: lshell is a Python-based restricted shell that limits users to a defined set of commands, enforces path and SSH transfer controls (SCP/SFTP/rsync), logs user activity, supports session/time restrictions, and more.

lshell is a Python-based restricted shell that limits users to a defined set of commands, enforces path and SSH transfer controls (SCP/SFTP/rsync), logs user activity, supports session/time restric...

via GitHub

GitHub - kvcache-ai/ktransformers: A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations

A Flexible Framework for Experiencing Heterogeneous LLM Inference/Fine-tune Optimizations - kvcache-ai/ktransformers

via GitHub

GitHub - headroomlabs-ai/headroom: Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server. - headrooml...

via GitHub

GitHub - JustVugg/colibri: Run GLM-5.2 (744B MoE) on a 25GB-RAM consumer machine — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦

Run GLM-5.2 (744B MoE) on a 25GB-RAM consumer machine — pure C, zero deps, experts streamed from disk. Tiny engine, immense model. 🐦 - JustVugg/colibri

via GitHub

Nous Psyche

Psyche is an open infrastructure that democratizes AI development by decentralizing training across underutilized hardware. Building on DisTrO and its predecessor DeMo, Psyche reduces data transfer by several orders of magnitude, making distributed training practical.

via GitHub

GitHub - pipecat-ai/pipecat: Open Source framework for voice and multimodal conversational AI

Open Source framework for voice and multimodal conversational AI - pipecat-ai/pipecat

via GitHub

65 - [Smartphone] Linux dans ton smartphone : où en est-on ?

Peut-on faire tourner Linux sur son téléphone ? Dans cet épisode technique, nous rentrons en détail dans le sujet : quels sont les obstacles techniques à surmonter, quels projets se penchent sur la question, comment s’organisent-ils entre eux, et qui les finance ? Pour y répondre, nous avons invité Arnaud Ferraris, fondateur de la distribution Mobian, qui nous dresse un état des lieux de ce qui est possible aujourd’hui, ce qui pourra l’être dans le futur, et les habitudes qu’il sera difficile de changer. En savoir plus : Pour vous accompagner dans l’écoute, on vous a concocté une retranscription complète avec des illustrations qui enrichissent le propos, et des liens pour aller plus loin Télécharger le fichier mp3 de l’épisode A lire ou écouter dans la même série : Smartphone Android et vie privée La coopérative Commown Supporter un téléphone Android sur le long terme - Fairphone /e/OS et Murena

via Podcast Projets Libres

AISI Blog | The AI Security Institute

View AISI research and work. The AI Security Institute is a directorate of the Department of Science, Innovation, and Technology that facilitates rigorous research to enable advanced AI governance.

via AI Security Institute

GitHub - open-jarvis/OpenJarvis: Personal AI, On Personal Devices

Personal AI, On Personal Devices. Contribute to open-jarvis/OpenJarvis development by creating an account on GitHub.

via GitHub

DIAGRAMMES

DIAGRAMMES

via DIAGRAMMES

Proxmox VE : bien débuter avec les conteneurs LXC - IT-Connect

IT-Connect Publié le 9 mars 2026 Par Florian BURNEL it-connect.fr Environ 16 minutes de lecture L’utilisation des conteneurs LXC sur Proxmox VE représente une méthode efficace pour exécuter plusieurs environnements Linux sur un même noyau, sans la lourdeur liée à l'utilisation de machines virtuel...

via Support francophone de Maxthon

L'esprit critique on Instagram: "S4E27 - La propagande #terrasse #dalle #béton"

119K likes, 2,082 comments - lesprit.critique on July 17, 2026: "S4E27 - La propagande #terrasse #dalle #béton".

via Instagram

GitHub - livekit/agents: A framework for building realtime voice AI agents 🤖🎙️📹

A framework for building realtime voice AI agents 🤖🎙️📹 - livekit/agents

via GitHub

Otto — Your agents. Always on.

A purpose-built device for OpenClaw. Deploy and manage always-on AI agents from your phone. No terminal required.

via Otto

Réglementation européenne : l'Eclipse Foundation outille gratuitement les développeurs face au CRA

L'ORC Learning Hub ouvre ses portes pour guider l'écosystème open source mondial face aux nouvelles exigences de cybersécurité de l'UE.

via Goodtech

Declare secrets once. Store them anywhere.

Declare what your application needs in secretspec.toml — resolve the values through any of 13 providers. No secrets in git. Same code, every environment.

via SecretSpec

GitHub - C4illin/ConvertX: 💾 Self-hosted online file converter. Supports 1000+ formats ⚙️

💾 Self-hosted online file converter. Supports 1000+ formats ⚙️ - C4illin/ConvertX

via GitHub

GitHub - NVIDIA/OpenShell: OpenShell is the safe, private runtime for autonomous AI agents.

OpenShell is the safe, private runtime for autonomous AI agents. - NVIDIA/OpenShell

via GitHub

GitHub - anthropics/claude-code: Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflo...

via GitHub