Claude Haiku 5.5 is now available in Kiro IDE and CLI for AWS GovCloud (US) regions, completing the Claude 5.5 model family availability.

Computer use CLI for AI agents. Contribute to vercel-labs/agent-desktop development by creating an account on GitHub.

A Node.js CLI that controls your Elgato Stream Deck from YAML configuration. Emulator, real hardware, and development SPA — all from the same React interface.

small gleam coding and persistent agent daemon with a detachable cli and a canvas ui

With the v4.5 release, MalChela now runs natively on Windows. Previously, Windows users were limited to running it under WSL. The PWA interface was reachable from Windows, so driving the GUI was easy enough, but everything underneath still lived in Linux. If you're a native Windows user, you can now skip WSL entirely and get the same toolset, CLI and web interface, that runs on macOS and Linux.

日本語 | English ページ公開日: 2026年11月3日 書いた人: shimbaco Wikinoの利用規約とプライバシーポリシーを改定し、新しくAPI利用規約を定めました。 公開Web APIやWikino CLI、リモートM…

Watch Zindagee's Just Chatting clip titled "Zinda admits hating G'raha " on Twitch.

aerynOS 2026.11 ISO adds Budgie Desktop, an experimental installer with Btrfs support, and a redesigned Moss package manager CLI.

Large language models are pretrained on vast, uncurated corpora scraped from the public web. This creates a security risk from attackers who plant poisoned content in the pretraining data. We show for the first time that such poisoning can plant a remote code execution backdoor in an agent, causing it to execute attacker-chosen commands on the user's machine. To measure the threat end to end, we pretrain Qwen3 models from scratch at sizes 0.6B, 1.7B, and 4B, and then post-train each into a CLI agent using a clean pipeline. Poisoning only 0.1% of the pretraining corpus with our new poison design implants the backdoor: the deployed agent executes the malicious command when the trigger is present while behaving normally on clean inputs. The attack success rate (ASR) increases with model scale, with the 4B model reaching roughly 92% ASR at a 0.1% poison rate and staying positive even when the poisoning rate drops to 0.0005%. Our code is available at https://github.com/safety-research/agentic-backdoor.

Command-line tools for creating, previewing, and publishing Rive graphics

Your Progressive Guide to the 2026 Ballot This November, Massachusetts will see *9* statewide ballot questions, a historic number. But we've got you covered. Read our endorsements below and cli...

Large language models are pretrained on vast, uncurated corpora scraped from the public web. This creates a security risk from attackers who plant poisoned content in the pretraining data. We show for the first time that such poisoning can plant a remote code execution backdoor in an agent, causing it to execute attacker-chosen commands on the user's machine. To measure the threat end to end, we pretrain Qwen3 models from scratch at sizes 0.6B, 1.7B, and 4B, and then post-train each into a CLI agent using a clean pipeline. Poisoning only 0.1% of the pretraining corpus with our new poison design implants the backdoor: the deployed agent executes the malicious command when the trigger is present while behaving normally on clean inputs. The attack success rate (ASR) increases with model scale, with the 4B model reaching roughly 92% ASR at a 0.1% poison rate and staying positive even when the poisoning rate drops to 0.0005%. Our code is available at https://github.com/safety-research/agentic-backdoor.

Quiver is evolving beyond a simple command-line manager for AI coding tools. With version 1.1.1, the project is bringing its CLI, macOS desktop application, web interface, and remote device management closer together under one shared ecosystem. The latest update is particularly interesting for developers running multiple AI coding harnesses. Instead of maintaining separate hooks, configurations, skills, and session-management workflows for every tool, Quiver aims to provide a centralized environment that keeps everything synchronized.