Full-text search across 69219 articles. Combine with topic and date filters; results sorted by relevance.
Google announced Gemini 4 Argon on September 30, 2026, a model it says leads on coding, agent and long-video tests, with initial access limited to trusted cyber defenders in Google's Fairwind Program before paid API cust…
Google Cloud AI Research released RRSI, an open-source framework for Regularized Recursive Self-Improvement of Agent Harnesses, on September 21, 2026, in collaboration with UNC-Chapel Hill, Stanford, and Washington Unive…
The Dutch Institute for Vulnerability Disclosure (DIVD) disclosed on September 24, 2026 that it was breached through its own infrastructure by an agentic AI-powered attack, and on September 30 identified the vector as tw…
Google restricted live access to its Gemini 4 Argon foundation model to vetted enterprise organizations and cybersecurity defenders under the Fairwind Program, citing the model's proficiency in reverse-engineering softwa…
C1 launched C1 LLM Gateway on Day 4 of its Launch Week, a policy-controlled endpoint that routes inference across supported public, private, and customer-controlled model deployments. The gateway ties each request to a p…
A developer running a daily autonomous Claude Code pipeline on a Windows PC described how two chained agents — a card writer that picks blog topics and a blog agent that publishes them — produced an ad placed under an un…
Ansight released version 1 of the Ansight CLI, a free tool that connects coding agents such as Codex, Claude and Cursor to a running iOS or Android app and records development and test sessions locally. The CLI, local ho…
A developer released astro-git-dates, an Astro plugin that sets a content note's date from its last git commit at build time, replacing hand-edited dates. The plugin wraps a content collection's loader and swaps the `git…
DepthFirst launched an AI-native continuous cyber defense product that scans code, dependencies, infrastructure and live environments, and published a benchmark comparing 34 model and thinking-level configurations on det…
WattzGOAT, an intentionally vulnerable web application built as a security-training lab, ships as a single Docker container with 48 hidden weaknesses to find and exploit. The made-up smart-meter company's customer portal…
A departing employee documented an offboarding method that uses an AI harness pointed at existing workplace connectors — GitHub, Jira, Confluence, Gmail, Drive, Calendar, Slack and the HR directory — to compile a persona…
A developer replaced the Rust `clap` crate with LLM-generated argument-parsing code written by Gemini 3.8 Flash via OpenRouter, cutting the compiled executable size by 300 kB and argument parsing time from 20 microsecond…
SemOPT, a semantic-guided framework for correcting LLM-based optimization models, achieves an average 7.6% accuracy improvement over the strongest baseline on complex datasets across seven optimization modeling benchmark…
The open-source AgentRQ task board and MCP server for Claude Code and other agents now records every task status change as an immutable, append-only history record instead of overwriting a single status field, via pull r…
Researchers submitted TokenCast, a method that learns a composable cost representation for each execution segment of an LLM agent run, forecasting token consumption before and during execution without additional LLM call…
Tim's Markdown Reader, a free open-source Markdown viewer for macOS released under the MIT license by its developer, renders files locally with no network requests, accounts, or telemetry. The Swift app requires Apple si…
BoxLite Cloud published a demo in which a coding agent built and deployed a playable mobile arcade game, Strike 1945, to a public URL from a single prompt, using the managed cloud for AI agents. The agent read the agent.…
A developer argues that git worktrees, while ideal for isolating parallel AI coding agent sessions within a single repository, hit a hard structural boundary in polyrepo setups because the shared object store is scoped t…
A developer tool called Ivar reduces Figma MCP configuration across multiple repositories to a single step, storing one MCP server definition in an ivar.json manifest that syncs into every agent session. Because Figma re…
Anthropic confirmed at 11:28 a.m. ET on Sept. 29 that its AI tools were returning to normal after an hour-long outage that began around 10 a.m. ET and affected Claude, Claude Code, Cowork, Console and API tools, with Dow…