---
title: 'In the News: October 29, 2026 (Extra 2)'
description: 'Google is winding down its Jules coding agent on December 31, with sessions ending November 11. Settings, memories and session history can be exported.'
canonical_url: 'https://darkfactory.dev/news/2026-10-29-extra-2'
markdown_url: 'https://darkfactory.dev/news/2026-10-29-extra-2.md'
collection: news
date_published: '2026-10-29T16:10:00-04:00'
date_modified: '2026-10-29T16:10:00-04:00'
---

# In the News: October 29, 2026 (Extra 2)


Google is shutting down Jules, its asynchronous coding agent, and anyone with work in it has until November 11. Separately, Meta published production figures for a system that generates tests designed to fail on risky code changes.

## 1. Google will wind down Jules on December 31, and sessions end November 11

**[Jules Turndown](https://jules.google/docs/changelog/2026-10-29/)** · Google · Jules changelog, October 29, 2026

Google's notice says Jules will wind down on December 31, 2026, "as part of our work to unify our agentic coding efforts." Users have until November 11 to finish sessions, after which Jules becomes read-only through the end of the year. Google describes Jules as an experiment in how asynchronous coding agents could "tackle your backlog," and says what it learned is shaping products across Google, including Google Antigravity. The notice does not say that Antigravity replaces Jules.

Google is offering an export through its Jules Tools command line package. The export covers account settings, environment variables, scheduled tasks, connected repositories, memories, setup scripts, and every session with its prompt, plan, activity log and code patches. Code, branches and pull requests stay in GitHub. The notice warns that the environment variable file holds the values themselves, so the export folder should stay out of your repositories. We read the notice but did not run the export.

**Why it matters:** If a pipeline of yours hands backlog work to Jules, the setup scripts, test commands and memories that shaped its output live inside the service, and the export is the only route Google documents for getting them out. That leaves 13 days before sessions stop. The same configuration is worth keeping as files in your own repository, whichever agent runs next.

## 2. Meta reports a 60% true-positive rate for LLM-generated tests built to fail on risky diffs

**[Catching Bugs Before They Land: JiTTesting's Human-on-the-Loop Approach at Full Scale](https://engineering.fb.com/2026/10/29/security/catching-bugs-before-they-land-jittestings-human-on-the-loop-approach-at-full-scale/)** · Sreeja Thummala, Sophie Zeng, Mark Harman, Weini Liu and Eric Liu, Meta · Engineering at Meta, October 29, 2026

Meta says its Catching Just-in-Time Tests system is now part of its infrastructure. It runs overnight on spare capacity against the previous day's diffs that a risk model flags. One workflow infers a change's intent and builds tests aimed at the ways that intent could go wrong. A second treats every diff as a bug and generates tests that expose any behavioral difference. Tests that fail on the change are candidates, which an ensemble of LLM judges and a rule-based filter called RubFake rank before anyone is contacted.

Meta's figures: 589 engineers, 64% of those contacted, engaged with the results, and they reported a 60% rate of true positives. In the first half of 2026 the system played "a key role" in catching 74 serious privacy bugs at the diff stage, and Meta says privacy incidents of that severity fell 20% over the same period. Meta calls the review model human-on-the-loop: engineers set criteria and "review outcomes at an aggregate level."

The page text we read does not say how the 20% figure was derived or what it was compared against. The 60% rate is engineer-reported, and we did not open the accompanying paper.

**Why it matters:** This is a verification stage built around executable evidence: a test that fails on the diff, rather than a model's opinion of it. Meta's pipeline gives two of its five stages to filtering weak catches before a person sees them. If you add a similar check to an agent loop, the filter and the reporting threshold are the parts to design first.