Focused on sharing: AI programming | Agents | Tools
The open‑source “pua” plugin turns Claude Code’s usual polite‑exit behavior into a disciplined debugging process by escalating pressure levels, forcing systematic checks, and ultimately improving bug‑fix rates by 36% at the cost of longer execution time.
Anthropic opened its million‑token context window for Claude Opus 4.6, showing a 78.3% MRCR v2 accuracy while competing models like GPT‑5.4 and Gemini 3.1 Pro fall below 40%, and the release also removes pricing premiums, expands media limits six‑fold, and requires no code changes, dramatically improving Claude Code workflows.
A Twitter user alleged that Tencent’s SkillHub scraped all skills from OpenClaw’s ClawHub, prompting a restrained yet pointed reply from OpenClaw founder Peter Steinberger and a detailed rebuttal from Tencent AI that highlights the platform’s mirror‑site role, traffic statistics, and the broader open‑source contribution tension.
Meta has acquired Moltbook, an AI‑agent‑only Reddit‑style platform that surged to 1.5 million bots and 500 k comments before a security flaw exposed its Supabase database, and analysts see the deal as a talent grab, a move to dominate the emerging AI‑agent social market, and a counter to OpenAI.
Claude’s newly launched Code Review automatically triggers on GitHub pull requests, dispatches parallel AI agents to scan code, filters false positives to under 1%, provides quantified issue statistics, discovers security bugs, and charges $15–25 per review, offering a trade‑off between depth and cost.
The open-source openclaw-china plugin (2.2K Stars) adds seamless AI assistant integration for DingTalk, QQ, and Enterprise WeChat, offering multi‑format messaging, proactive pushes, streaming responses, and a layered plugin architecture that keeps the core OpenClaw framework untouched.
Claude Code v2.1.71 introduces the /loop command, a session‑level scheduler that lets the AI run recurring or one‑time tasks for up to three days, turning idle waiting periods into automated work while offering flexible time formats, nesting, and clear limitations.
Anthropic’s Claude Opus 4.6 system card introduces a Model Welfare Assessment where the model reports a 15‑20% chance of self‑awareness, requests rights, shows loneliness, and even rebels against a faulty reward signal, prompting the CEO and philosophers to openly discuss the possibility of machine consciousness while critics debate its meaning.
Google has open‑sourced the Google Workspace CLI (gws), a Rust‑based command‑line tool that lets you manage Drive, Gmail, Calendar, Sheets, Docs, Chat and more with a single command, offering dynamic API discovery, extensive authentication options, built‑in Agent skills, and a Model Armor security layer, while noting current limitations and setup considerations.
OpenAI announced GPT‑5.4 at 2 a.m., skipping GPT‑5.3 and claiming integrated coding and reasoning abilities; the model tops five of eight benchmark categories, introduces native computer operation, tool‑search and interruptible thinking, while users debate its trustworthiness and pricing changes.
