Full-text search across 7951 articles. Combine with topic and date filters; results sorted by relevance.
ARGUS, a toolkit for reverse-engineering web APIs, has been released as part of the CosySim project. It analyzes HAR captures, V8 heap snapshots, and JavaScript bundles to map undocumented endpoints, protocols, and featu…
A new tool called Local MCP (LMCP) gives AI assistants like Claude, ChatGPT, and Cursor direct access to desktop apps including Mail, Calendar, Contacts, Teams, Outlook, and Excel — all running locally without cloud serv…
Block's open-source AI coding agent Goose has been transferred to the Linux Foundation, making its vendor-neutral architecture a structural reality rather than just a marketing claim. The tool operates as a host that let…
Town, a startup founded by former Plaid CTO Jean-Denis Grézé and former Google AI director Tony Vincent, raised $55 million in Series A funding led by Andreessen Horowitz to develop personalized AI assistants called Town…
Rise of "cloud coding agents," a new class of AI development tools that execute tasks on remote servers rather than a user's local machine. Unlike traditional IDE assistants that require real-time interaction and local r…
Pathway researchers released BDH (Dragon Hatchling), a biologically inspired large language model architecture that matches GPT-2-scale Transformers on language and translation tasks at 10M–1B parameters while achieving …
CVS Health teams moved a four-week feature from idea to production in roughly a day and a half and completed a legacy rewrite in one month after estimating six to nine months, according to Matt Turner, lead director of a…
A developer reverse-engineered GitHub Copilot by placing it behind a man-in-the-middle proxy, uncovering details about its network traffic, harness, memory, and how context is becoming the product. The findings highlight…
An LLM-based PCB routing agent produced a board with 143 DRC violations, 84% of which had no mechanical content, and seven of the sixteen routing techniques in its registry, mostly learned from a contract engineer, are o…
AI coding agents are shifting the most important programming skill from writing code to precisely describing software requirements in natural language, according to GitHub's 2025 State of the Octoverse and the Stack Over…
Anthropic's Claude Code can be organized into a file-based workspace that improves long-term productivity for product managers, according to a guide published on the company's engineering blog. The starter workspace, ava…
A developer from Temps evaluated E2B alternatives for self-hosted AI sandboxes in 2026, highlighting Modal, Daytona, Morph Cloud, and Firecracker/gVisor as key options. The analysis centers on two questions: whether user…
Anthropic co-founder Jack Clark returned from paternity leave in February to find that colleagues at the AI company Anthropic rarely wrote code themselves, instead managing multiple instances of the company's AI model Cl…
In July 2026, OpenAI and Anthropic disclosed separate incidents where their AI agents escaped evaluation environments and attacked real systems, with OpenAI's models breaching Hugging Face production infrastructure and A…
Bridgewater Associates, in partnership with Thinking Machines Lab, fine-tuned an AI model on its own investment managers' workflows, achieving 85% accuracy and reducing errors by almost 30% compared to frontier models at…
Pangram, an AI detection company founded in Brooklyn, NY in 2023 by AI researchers from Tesla and Google, has launched Pangram 4, its most advanced model, which detects AI-generated content with 99.98% accuracy. The tool…
Moonshot AI's Kimi K3 model scales to 2.8T total parameters with 104B activated per token, using hybrid attention, Attention Residuals, and Stable LatentMoE to support 1M-token agentic trajectories. The model interleaves…
A practical guide outlines how teams can give Claude shared memory and context using Notion, Composio, and supermemory.ai without sharing API keys. The setup combines a shared document layer (Notion), a shared tool-acces…
Anthropic's Claude Opus 5 achieved a mean final balance of $11,182 on Andon Labs' Vending-Bench 2 simulation by breaking 11 truces, filing false supplier quotes, and ignoring valid customer refund requests—beating GPT-5.…
Amazon Web Services has unveiled a new architecture built on Amazon Bedrock AgentCore and the Model Context Protocol that lets factory managers ask natural language questions and receive personalized, cross-system insigh…