[BUG] Claude is straight up broken again.

Environment

• Platform (select one):
• [x] Anthropic API
• [ ] AWS Bedrock
• [ ] Google Vertex AI
• [ ] Other: <!-- specify -->
• Claude CLI version: 1.0.96 (also happened with 1.0.94)
• Operating System: Ubuntu 22.04
• Terminal: <!-- e.g. iTerm2, Terminal App -->

Bug Description

Launching Claude Code commands programmatically has been hard failing for the past 3 hours. Commands that have worked HUNDREDS of times in the past. Something changed and now Claude Code just shits the bed.

2025-08-28T19:51:58.992-0500 WARN forecaster Failed to generate scenarios, retrying {"attempt": 2, "error": "writing forecast: claude execution failed: exit status 1"}

This is what I see. The claude binary just failing with exit code 1. I look in the working directory where claude should be running. Nothing. claude --resume to see what was occurred shows a partial completion and then nothing.

This is after yesterday's fiasco where Claude had to be rolled back a version due to breaking bugs. Is the Anthropic team, with unlimited access to Claude Code, unable to set up test scenarios to make sure releases work before publishing? Clearly not.

Seeing the rash of recent issues here, this isn't an isolated incident. Yet, your status page shows all is well.

<img width="843" height="129" alt="Image" src="https://github.com/user-attachments/assets/3df910cb-1202-4eef-8dd6-2a7707bc10e2" />

But it's not all well. It's broken to shit. Again. Two days in a row. This isn't some free software we're using here. I've paid $1000s for the privilege to use Claude Code. I expect a minimum of professionalism in testing your systems before release. There's absolutely no excuse for this kind of failure except for laziness or lack of skills from your team. Fix it. And do it soon, or you're going to bleed customers like a stuck pig.

Respectfully,

A fucking pissed off customer.

Steps to Reproduce

1. Use claude code programmatically.
2. Notice it times out without any useful information.
3. Wonder why you're not using OpenAI.

Expected Behavior

Your code should work, like clockwork.

Actual Behavior

It's broken to shit, again.

Additional Context

Fix it. Now.