Skip to content
Tech News
clear
Topics: Today This Week This Month This Year

Community fork 'Maiao' revives Gerrit-style stacked PR tool after Adevinta abandons it

Maiao, a command-line tool that recreates Gerrit's stacked-commit review workflow on modern git hosts, has been picked up by community maintainers under runetes/maiao after Adevinta's original team stopped maintaining it. The tool's 'git review' command turns each commit in a branch into its own linked pull or merge request, automatically managing parent-child dependencies and rebasing the stack as changes get merged. It now supports GitHub, GitLab, Gitea, Forgejo/Codeberg, Bitbucket Cloud, and the beta Cursor Origin platform, auto-detecting the provider from the repository's remote URL.

OpenAI launches Admin plugin for ChatGPT Work and Codex

OpenAI has released an Admin plugin for ChatGPT Work and Codex that lets workspace administrators manage users, permissions and spending limits through a chat interface rather than a separate dashboard. The tool can generate usage reports, flag credit thresholds, and handle onboarding or offboarding tasks on request. A company demo showed an admin querying adoption trends and requesting visual breakdowns for a monthly review.

Life360 launches $7.99 QR pet tags with app-based lost-pet alerts

Life360 is rolling out a new line of scannable pet tags priced at $7.99, expanding beyond its GPS-based pet trackers introduced last year. Scanning the tag's QR code sends contact details to whoever finds the pet and simultaneously notifies the owning family through the Life360 app, without requiring the finder to have the app installed. The company is also adding pet care tools, including scheduled care logs, to its platform.

Developer runs Qwen 35B model on Raspberry Pi 5 for offline car assistant CarWatch

A hobbyist built CarWatch, a system that puts a Raspberry Pi 5 in a car running a 35-billion-parameter Qwen model locally, no internet or subscription required. It joins chat rooms as an AI agent, sending trip summaries, dashcam clips, and answers sourced from the car's 745-page owner's manual, and also supports hands-free voice queries via on-device speech recognition.

IBM's Harvest Ran the NSA's Cold War Code-Breaking for 14 Years

IBM built a top-secret cryptographic computer called Harvest for the National Security Agency, which operated from the Cuban Missile Crisis in 1962 until the mid-1970s. The machine processed data up to 200 times faster than other computers of its era and included the first automated tape library, robotically retrieving magnetic tape cassettes on demand. IBM also created a custom programming language, Alpha, so analysts could describe cryptographic problems the way Fortran was used for scientific computation.

OpenAI's Thibault Sottiaux details strategy behind ChatGPT Work rollout

In an interview, OpenAI product lead Thibault Sottiaux explained that ChatGPT Work is meant to extend the capabilities of its Codex coding agent to non-technical, white-collar users. He said the product was bundled into the existing $20 Plus plan to make AI agent tools accessible to as broad an audience as possible, framing the effort as repackaging technical agent infrastructure into something safer and simpler for everyday use on mobile and web.

OpenAI reinstates 5-hour usage cap for Codex and ChatGPT Work on Plus tier

OpenAI will restore a five-hour usage limit for Codex and ChatGPT Work on Plus subscriptions starting tomorrow, according to engineering lead Thibault Sottiaux. The cap had been temporarily suspended for several weeks, leaving only a weekly limit in place, but is now returning alongside that weekly cap. Pro subscribers on the $100 and $200 tiers will remain exempt from the five-hour restriction for the coming months.

Developer builds Claude Code skill to recover Amazon's blocked Kindle highlights

A developer released a Claude Code plugin skill that pulls every highlight from a Kindle book's notebook page, including ones Amazon's export feature truncates or hides entirely due to an undocumented export cap. It reconstructs the missing text by cross-referencing the Mac Kindle app's synced annotation positions and the Cloud Reader's rendered pages, producing one combined, verbatim Markdown file with location citations.

vLLM tool-parser bug let AI-generated text trigger code execution

Researchers documented that vLLM's XML-based tool parser for Qwen3 Coder (CVE-2025-9141) fed model-generated arguments directly into Python's eval() function, meaning a language model's own output tokens could execute arbitrary code on the machine running it. A Gemini-based review flagged the change as a critical risk before it was merged, but the lead vLLM maintainer force-merged it anyway.

Today's top topics: android authority iphone duo foldable phone meta chatgpt elon musk
Browse all topics →  ·  Today's trending topics →