---
title: 'In the News: November 13, 2026 (Midday)'
description: 'Claude Code 2.1.323 removed subagentStatusLine, and a contributor now says a grace period is coming. Allspice reports its hardware agent beats Claude Code and Codex.'
canonical_url: 'https://darkfactory.dev/news/2026-11-13-midday'
markdown_url: 'https://darkfactory.dev/news/2026-11-13-midday.md'
collection: news
date_published: '2026-11-13T13:25:00-05:00'
date_modified: '2026-11-13T13:25:00-05:00'
---

# In the News: November 13, 2026 (Midday)


Claude Code 2.1.323 removed the setting that drew per-subagent cost, context and model rows, and the people who built on it have no replacement yet. A contributor to the repository now says a grace period is coming in the next release. Separately, Allspice describes a hardware design agent that edits schematics through a narrow tool interface and reports that it beats Claude Code and Codex on its own evals.

## 1. Claude Code 2.1.323 removed subagentStatusLine, and a contributor says the removal needed a grace period

**[2.1.323 removed subagentStatusLine with no replacement: per-agent context, cost, model and effort are now invisible](https://github.com/anthropics/claude-code/issues/109446)** · rezzminator and other users · GitHub issue #109446, anthropics/claude-code, opened November 13, 2026 (UTC)

The v2.1.323 release notes, published November 12, list "Removed the subagentStatusLine setting". Subagent rows now use the default look, and settings files that still hold the key load without error. The issue's author reports that 2.1.322 ran the configured command and 2.1.323 never calls it, with no startup warning. By the users' account, the custom rows showed each running subagent's context use, token and dollar spend, model, reasoning effort and prompt-cache state. They also say mods cannot redraw those rows, because no mod render site covers the agent panel, and that a request for one filed October 29 has had no answer. The thread had 22 thumbs-up when we read it.

A contributor who said he made the change replied that plugins shipping a settings file made little sense with only two supported keys, and that he should have given a grace period. He said a grace period for the key will go in the next release, that the replacement is planned for mods, which have not landed yet, and he gave no ship date. One commenter's stopgap is to pin 2.1.322 with auto-update off, which gives up every later fix. The statusline documentation page no longer mentions the setting when we read it, although the issue says it did.

**Why it matters:** If you built per-agent cost or context displays on this setting, 2.1.323 broke them, and what is on offer is a temporary grace period, not a restoration. Until that release ships, the choices in the thread are pinning 2.1.322 or going without the rows. We did not check whether the grace period has shipped.

## 2. Allspice reports that its hardware design agent beats Claude Code and Codex when given schematic-editing tools

**[How Allspice built a frontier hardware design agent](https://allspice.io/post/how-allspice-built-a-frontier-hardware-design-agent)** · Hayder Tirmazi, Principal Software Engineer, Allspice · Allspice blog, updated November 13, 2026

Allspice built DRCY, an agent that reviews hardware schematics and, in its new version, also modifies and creates them. The first version was a fixed sequence of specialized models; the new one uses a harness written in Rust that runs server-side in sandboxed containers. Tirmazi says coding-agent habits did not carry over, because a schematic is not a text file an agent can edit and run. DRCY's harness instead gives the agent a Hardware Mutations Interface, a set of well-scoped tools that change a schematic without the agent parsing its underlying ECAD format.

Allspice says this lets the agent create correct designs at a higher success rate than off-the-shelf harnesses using the same underlying model. The comparison used Claude Code 2.1.322 with Claude Opus 5.5 and Codex CLI 0.167.0 with GPT 6.1 Sol, on a randomly sampled subset of Allspice's own human-verified tasks that the company says was held out from training and tuning. The off-the-shelf runs did not have DRCY's mutation or platform tools. In one example, drawing a voltage divider from two 10 kΩ resistors, the post says Claude Code placed it past the sheet's right border and Codex drew it without resistors.

The success rates are in charts, and we did not read the values. The evals are internal and the author is the vendor. Because the comparison removes DRCY's tools from the others, it measures the tool interface and harness together, not the model.

**Why it matters:** For anyone building an agent over something that is not plain text, the post's claim is that a small intent-level tool layer beats handing a general harness the raw files. That claim rests on one vendor's internal eval, so treat it as a design to test against your own tasks.