Full-text search across 58710 articles. Combine with topic and date filters; results sorted by relevance.
Four major surveys converge on the finding that AI-generated code fails in production 43% of the time, according to byteiota. The report says AI-generated code is faster to write but harder to debug, and it presents the …
An open-source developer released Wy, a Rust terminal tool that surfaces the agent-session context behind AI-generated code changes, aiming to close the gap between what a git diff shows and why an agent made the change.…
DigitalOcean launched Managed Agents, a service that runs Claude Code and Codex CLI inside Firecracker microVMs with active-CPU billing at $0.044 per hour and access to 16,000 managed tools. The offering targets develope…
A third-party GitHub Action called PR Comprehension Quiz Gate requires pull request authors to answer an AI-generated quiz about their own code changes before merging. The Action uses Copilot or Claude to write and grade…
A developer released a GitHub Action, yuyuyuyuyu-dev/assert-no-unexpected-changes, that runs an AI-written code change in two Docker containers — one executing only the `arrange` setup phase and one running both `arrange…
Island researchers Oleg Zaytsev and Ofek Ronen disclosed a human-operated phishing platform that impersonates advertising products for Google Gemini, Anthropic Claude, OpenAI ChatGPT, Perplexity, Meta Muse, and Manus, us…
Phishing campaigns impersonating ChatGPT, Gemini, Claude, Perplexity, and Meta Muse are stealing passwords, multi-factor authentication codes, and access to advertising accounts, according to CyberUpdates365. The fake AI…
Orbtile launched on Hacker News' Show HN as a macOS app that turns a Logitech MX Creative Keypad into a live status display for AI coding agents, showing one key per session with state, tool name and latest text. The app…
Reflection AI has built Beam, a 501-billion-parameter open-weight model for coding and agent tasks, and plans to publish the weights under an Apache 2.0 license later this month. Beam uses a mixture-of-experts design in …
Joshua built Homestead, an open framework around Claude Code that gives each AI agent a shared channel to see and message other agents, routes incoming notifications to project-specific agents, and pairs a Worker with a …
DeepSeek released Harness Desktop v0.2, a GUI installer for its open-source coding agent Harness that removes the Node.js and terminal prerequisites previously required to run the CLI tool, which launched in August 2026.…
A developer released "What's Agent Doing," an MIT-licensed Claude Code UI mod that displays the current step, a timer, and per-background-agent status in a single line above the prompt. The plugin uses Claude Code's new …
A developer reported that Claude Sonnet 5.5 generated two Rust functions for about $0.06 after DeepSeek V4.1 Flash produced hard-to-read code and a slight performance regression. The developer said Sonnet fixed the regre…
A project called Llmtobinary lets large language models generate machine code directly, byte by byte, for ANMA-1, a processor designed to be programmed by AI without source code or a compiler. Users describe a program, w…
MonoCode launched a downloadable Mac GUI that wraps coding agents including Claude Code, Codex, Cursor, Grok Build, OpenCode, Antigravity, Pi, omp, fx, and Hermes Agent. The tool is distributed via a Mac download and a G…
A developer released tokenmaxxing, an open-source CLI and public leaderboard that aggregates AI coding-agent usage across multiple tools and machines into a single profile. Built on ccusage, the tool syncs only daily agg…
A computer repair shop owner who builds apps on the side created Zeltro, a desktop app that gives vibe-coding projects a single home with per-project addresses, a shared database, and an embedded terminal where AI agents…
ProgramDistill, a benchmark for real-world app reconstruction tasks, found that AI coding agents fail 51% of the time, including GPT-6 Astra, according to byteiota. The result highlights a persistent gap between AI codin…
Jindo AI is building Pawsly, a coordination layer for multi-agent coding workflows that shares plans, roles, decisions and commit state across parallel agent sessions. In internal benchmarks on ten two-agent coding tasks…
Researchers developed EMBODIEDSWE-BENCH, a simulation benchmark for coding agents that spans contact-rich manipulation and deformable object tasks in long-horizon, dexterous robotics, to test whether agent-generated solu…