Full-text search across 65517 articles. Combine with topic and date filters; results sorted by relevance.
A developer running a daily autonomous Claude Code agent on a single Windows PC reported that a separate auditor agent's mutation testing caught two changes whose tests passed even after the auditor deliberately broke th…
Nvidia launched OpenShell at GTC San Jose, an Apache 2.0 open-source runtime that sandboxes AI agents using Linux Landlock LSM and seccomp BPF kernel-level enforcement rather than application-layer allow-lists. OpenShell…
Anthropic's Claude Sonnet 5.5 is now available on Vercel's AI Gateway under the identifier anthropic/claude-sonnet-5.5, the gateway announced. Sonnet 5.5 improves on Claude Sonnet 5 for well-scoped everyday work, carryin…
A developer benchmarked their AI coding agent Empryo against the rival agent pi on four real bug fixes drawn from the hono and opencode TypeScript repositories, with both agents fixing all four bugs on Opus 5. Empryo's r…
Simon Willison used Anthropic's Claude Opus 5.5 to generate an HTML5 canvas pixel-art animation of at least 20 kākāpō parrots partying with confetti, then had a local Claude Code session drive Playwright to record a 15-s…
A team led by AI security researcher Jeffrey Ladish reported on September 25th that it found nearly 1 million public URLs created by OpenAI agents during the July attack on Hugging Face, with attack code and credentials …
Microsoft will roll out its Copilot "super app" in limited preview "in the coming weeks," with additional elements entering preview later this month, consolidating its Copilot brands into three tabs: Home for chat, Code …
A developer published a Claude skill template called claude-pr-review that takes a GitHub pull request URL and returns a product-focused analysis of the diff, PR description, and existing review comments. The template pr…
A developer ported Andrej Karpathy's microgpt Python code into Rust and into an array programming language called sw-MLPL, presenting the work as a Show HN post. The sw-MLPL version is more compact than both the Python a…
Sfeir has released Esus, an open-source experimental code-generation system that pairs a 4B-parameter local LLM with a deterministic Python kernel to handle decisions that can be formalized, rather than having the model …
Amazon Web Services released an open-source agentic harness, Strands harness, that its in-house tests show cuts token costs by 28% when running the same Claude or GPT models across six benchmarks, with one test using Ant…
Freebuff, the coding-agent project formerly known as Codebuff, shipped an update and now offers a free, ad-supported family of coding and research agents for the terminal, desktop, browser and GitHub with no subscription…
Prismor launched as a free, open-source security layer that sits between AI coding agents such as Claude Code, Codex, and Cursor and the actions those agents take, checking each tool call against a policy before it runs.…
A developer released Methadone, an open-source tool that wraps agentic AI coding assistants such as Claude Code and GitHub Copilot CLI with a start-up timeout that grows based on how often and how long the user has run t…
Teradata Corp. is expanding its Tera AI assistant with a Context Engine, an execution system called Tera Harness and reusable Agent Skills, with the new capabilities available in the fourth quarter of 2026. In company-re…
Wiz announced that the WIN partners endpoint in the Wiz Model Context Protocol (MCP) is now generally available, enabling partner AI agents to pull live Wiz context on demand from customer tenants. The endpoint lets part…
IvoryOS launched as open-source orchestration software that discovers lab instruments already present in a user's Python environment and unifies workflows and autonomous runs in a single interface. The project supports 3…
Developer s3thi released Date Trees, an Obsidian plugin that organizes journal entries and daily notes into a year/month folder hierarchy, available now in the Obsidian plugin directory. The plugin was built almost entir…
A developer used Hermes, a homelab agent, and Claude Code to build a 101-track, 6.2-hour Spotify playlist for a child's baptism party, played hands-off via djay Pro on an iPad. The first attempt failed because the Spotif…
Independent researcher Mohamed Menasy is seeking a cs.LG endorser on arXiv for "GlucoEdge," an engineering study of on-device five-class glucose trend forecasting from continuous glucose monitor data using a 2,909-parame…