CATCH #1921
DETECTED OCT 8, 2026 05:25 UTC
THE CATCHscore 5/10
EXHIBIT 1921
Oct 8, 2026 05:25 UTC
github_releases+pages
anthropics/claude-code
Claude Code (anthropics/claude-code) released v2.1.294, improving `prompt` and `agent` hook instruction handling to prevent early stopping and correctly block commands.
READY TO POST:
looks like claude code just dropped v2.1.294, fixing some prompt and agent hook instruction issues. claude should be less likely to stop early now.
anthropics' claude-code got an update! v2.1.294 apparently makes prompt hooks better at judging instructions, meaning fewer premature stops.
EVIDENCE — the receipts
[MERGED across 2 sources: github_releases, pages]
shared signal: claude-code
--- github_releases :: anthropics/claude-code ---
anthropics/claude-code release: v2.1.294
name: v2.1.294
published: 2026-10-08T05:03:54Z
url: https://github.com/anthropics/claude-code/releases/tag/v2.1.294
## What's changed
- Fixed `prompt` and `agent` hooks written as instructions (such as "Block commands that...") allowing what they should block
- Improved how `prompt` hooks on Stop and SubagentStop written as instructions (such as "Carry on if the build is broken") are judged, so Claude is less likely to stop early
--- pages :: https://raw.githubusercontent.com/anthropics/claude-code/main/CHANGELOG.md ---
Claude Code changelog changed: https://raw.githubusercontent.com/anthropics/claude-code/main/CHANGELOG.md
3 added lines
---
+++
@@ -1,2 +1,5 @@
# Changelog
+## 2.1.294
+- Fixed `prompt` and `agent` hooks written as instructions (such as "Block commands that...") allowing what they should block
+- Improved how `prompt` hooks on Stop and SubagentStop written as instructions (such as "Carry on if the build is broken") are judged, so Claude is less likely to stop early
## 2.1.293
caught by the 20-minute sweep ·
the live wire · all receipts