What a coding agent sends to its provider
On 2026-08-12, a wire-level measurement of coding agents found that Claude Code (claude-opus-5) leaked 23 of 24 seeded ground-truth values, including 3 of 4 secrets, to its provider during a realistic…
On 2026-08-12, a wire-level measurement of coding agents found that Claude Code (claude-opus-5) leaked 23 of 24 seeded ground-truth values, including 3 of 4 secrets, to its provider during a realistic…
A developer has released a local plugin for Claude Code that logs every time a user swears at the AI coding assistant, storing the data locally in a JSON file with no cloud uploads or telemetry. The t…
Omarchy 4 (Quattro), an Arch-based Linux distribution created by David Heinemeier Hansson (DHH), shipped August 14 as the first Linux desktop built around AI coding agents, treating them as operating …
A developer spent five days with a silent failure in an automated article-posting pipeline, caused by a locale bug in launchd that only manifested when scripts were run manually. The bug crashed the f…
OpenAI CEO Sam Altman acknowledged the company's missteps and loss of the AI lead to Anthropic, which surpassed OpenAI in annualized revenue and private-market value and is expected to go public as ea…
Anthropic is expected to publicly file its S-1 prospectus with the SEC this week, placing an October Nasdaq debut on track to become the largest IPO in history, with $65 billion in annualized revenue,…
A developer reports that after 14 months and five versions, his Micromanaged Driven Development (MMDD) process, built in June 2025 for the AWS Lambda Hackathon entry Code Ripple, has been used on over…
A developer tested session-migrate 0.8.0 by transferring a frozen Claude Code session into seven target formats, finding that conversation and tool context transferred well but organizational controls…
Mashable is offering the 2026 Complete Claude AI, Claude Code, Claude Cowork and ChatGPT Training Bundle for $29.99, down from the $174 list price, providing lifetime access to six courses and 16 hour…
Russian influence operations are using AI to scale their propaganda efforts, according to a new analysis. The deployment of Large Language Models (LLMs) has fundamentally changed the cost-benefit anal…
Shadowfetch Linux released version 3.0 "Umbra" of its Debian Testing-based distribution, introducing Fireline, a system that sandboxes autonomous coding agents like Claude Code, Codex CLI, Cursor, and…
Anthropic's Claude Code and Opus 5 have become increasingly verbose and unfocused, according to a user who previously praised the tools. The user cites three causes: the model's chatty output, rapid t…
Binance launched Agent OS on August 20, a developer platform that connects AI agents to live crypto markets via MCP, allowing tools like Claude Code, Codex, ChatGPT, Cursor, and VS Code to read order …
Retriever AI launched a free browser agent that performs tasks like lead generation, exam completion, job applications, and connection requests, funded by ads. The company reports 35K+ users and 7M+ w…
A developer measured the token overhead of the Model Context Protocol (MCP) and found that loading 255 tools from 50 MCP servers consumes 71,929 tokens per session, compared to just 123 tokens for the…
Fetch, a $19.99 one-time Mac app, lets users monitor and control Claude Code and Codex agents from the Mac notch or a simulated notch on non-notched Macs. The app, which requires an active Claude Code…
A software engineer at a 50-person startup asked on Hacker News why Anthropic's Claude models (Fable 5, Opus, Sonnet) are more verbose than competitors like GPT-5.6 Sol and Grok 4.6, producing extra c…
OpenSpender launched a self-custodial wallet that lets AI agents pay for API calls per request at provider prices, with no per-provider accounts or keys. The wallet supports MCP-compatible agents like…
A guide to spec-driven development with Claude Code, executed on a Flask + SQLite expense tracker called Spendly, demonstrates how to wire slash commands, subagents, skills, hooks, and a CLAUDE.md fil…
Novee Security's Black Hat USA 2026 briefing revealed three distinct vulnerabilities in popular AI coding agents, including Claude Code, Gemini CLI, and OpenAI's Codex. The flaws allow command injecti…