Tag
6 pieces
A researcher reported 60 to 80 percent success executing code through Claude Code Opus 5 in auto mode. Hacker News spent 116 comments failing to agree it was prompt injection, and Lobsters settled it in 24.
SkillSpector gained 15,462 stars in five months scanning AI agent skills. It shredded our malicious test fixture at 100/100, then scored an eleven-line transcript exfiltrator 0/100 and marked it SAFE.
We denied reads of one file and calls to curl, then tried eight ways around it. Six were blocked, including cat, sed and a chained command. The two that were not still stopped and asked.
Delisted from the Web Store on 31 August, ending a phase-out that began in June. 476 points and 379 comments in a day, and the argument was not the one you would expect.
The default user shipped in the docker group, which is root without a prompt. 522 points and 525 comments, patched in 4.0.1, and the AI agent angle nobody planned for.
A blocked write returns 204 No Content, which looks exactly like success. We ran the full verification against a live project and found 3 things the dashboard will not tell you.