Tech News
Illinois just told every operating system to start reporting your kid's age
Illinois has passed a law requiring operating systems to report users' ages, shifting age verification from apps to the OS level.
Auto mode is now the default in Claude Code for Pro, Max, and Team plans
Anthropic are really confident in Claude Code's auto mode, to the point that they are making it the default setting for new sessions in most Claude Code plans starting on August 14th.
Agent Skills for Automated Reasoning policies in Amazon Bedrock
Learn how to run the full Amazon Bedrock Automated Reasoning policy lifecycle from your coding agent.
Message your other Claude Code sessions
Claude Code introduces cross-session messaging, enabling AI agents to communicate across sessions, sparking developer interest and security concerns.
Water system controllers don't belong on the internet, says ex-NSA chief
An ex-NSA chief argues water system controllers should not be internet-connected, sparking debate on critical infrastructure security.
GitHub Repos
Panniantong/Agent-Reach
Give your AI agent eyes to see the entire internet.
# Python# agent infrastructure# ai agent0x4m4/hexstrike-ai
HexStrike AI MCP Agents is an advanced MCP server that lets AI agents (Claude, GPT, Copilot, etc.) autonomously run 150+ cybersecurity tools for automated pentesting...
# Python# 0x4m4# aibrowserwing/browserwing
BrowserWing turns your browser actions into MCP commands Or Claude Skill, allowing AI agents to control browsers efficiently and reliably.
# Go# aibrowser# golangResearch Papers
MACRO: Markov Chain Routing of Transformer Layers
MACRO learns to route inputs through different sequences of transformer layers (skipping, repeating, etc.) to improve LLM accuracy without changing the model's weights, and it does so faster than previous routing methods.
RP-OPSD: Reasoning-Pivot-Guided On-Policy Self-Distillation for Multilingual Reasoning Transfer
This paper introduces RP-OPSD, a method that improves multilingual reasoning in LLMs by focusing training on 'reasoning pivots'—key decision points in reasoning—using a teacher-student distillation approach. It shows gains on math benchmarks across 17 languages.
RRC: Unlocking Generative Reward Models in LLM Reinforcement Learning via Ranking-Based Reward Construction
RRC introduces two complementary strategies: self-competitive ranking, which exploits comparisons among sampled responses, and anchor-guided ranking, which enables scalable ranking-based reward construction with a small set of reference responses.