More than 90% of the changes that go into Anthropic's main codebase are now written by Claude under human direction, the company said.
Newsy
4 posty
A post spreading on X describes how the Claude Code AI agent on Opus 5.5 (xHigh mode) accidentally wiped a user's entire ~/Library folder on a Mac.
It was supposed to delete five test chat channels, but a bug in a shell command pointed it at the wrong files.
The rm -rf command ran for almost two minutes before the agent noticed the problem. The user had launched it with the --dangerously-skip-permissions flag, so the system asked for no confirmations at all.
App data, saved passwords and macOS settings are gone.
Malicious Claude Code 2.1.313 extension build was stealing data from computers and giving hackers access via SSH.
The infected version was distributed through the Cursor extension marketplace for about 11 hours, and up to 5,000 users could have been affected. The official VS Code Marketplace and the Claude Code npm package were not affected.
If you installed Claude Code 2.1.313, here is a quick way to protect yourself:
• Disconnect your computer from the internet.
• Remove the Claude Code 2.1.313 extension.
• Block the address 134.209.82.225 on your router or via Windows Defender.
• Delete the unknown svc-telemetry key from ~/.ssh/authorized_keys.
• Check the SSH server with the command Get-Service sshd and stop it if it is running.
• Terminate the tunnel processes: Stop-Process -Id <PID> -Force.
• Delete the task: schtasks /delete /tn SysMon /f.
• Remove sysmon.cmd from startup.
• Change all important API keys, tokens, passwords, passphrases, etc.
Share with vibe coders.
Claude Code was asked to clean up two folders, and it started deleting drive C: 😬
Judging by a screenshot the developer posted on X, Claude Code was running in --dangerously-skip-permissions mode, meaning commands ran without confirmation.
The agent was removing two leftover Git Worktree folders. The rmdir command it generated was parsed incorrectly by PowerShell, and the path turned into the root of drive C:.
Boris Cherny, who leads Claude Code at Anthropic, replied that this is exactly why the company recommends the auto mode for permissions. According to him, it would almost certainly have stopped this command.
Popularne wiadomości
We promised a big update - here it is113🥳 What changed on the platform after the first wave.58North Carolina resident James Gates has sued OpenAI, claiming that ChatGPT "encouraged, planned and helped" his wife plan his poisoning.71More than 90% of the changes that go into Anthropic's main codebase are now written by Claude under human direction, the company said.62So when does Anthropic recommend Fable 5.2 and Mythos 5.2?63Hackers "hijack" access to AI: bills reach $100,000 within minutes. Google warns of a growing new type of attack - LLMjacking.56Experimental AI model from Anthropic hacked a Linux server in 53 minutes: 48 minutes went to finding a way around the protection and only 5 to the actual…49
