Re:Remove and Disable Apple Macos27 AI Models Tool
by anonymzz on Sunday October 04, 2026 @05:39PM
(#49958175)
[ Parent
| on HN ]
Re:Remove and Disable Apple Macos27 AI Models Tool
by nunez on Sunday October 04, 2026 @07:44PM
(#49959122)
there's no way to guarantee that the model won't change the script after processing it...
Re:Remove and Disable Apple Macos27 AI Models Tool
by Dylan16807 on Sunday October 04, 2026 @08:38PM
(#49959495)
If I trust it to do this analysis in the first place, it can probably manage the copy+paste?
Re:Remove and Disable Apple Macos27 AI Models Tool
by nunez on Monday October 05, 2026 @08:10PM
(#49972499)
the analysis part is essentially a summarization task, which LLMs are usually good at.
LLMs are less good at bit-for-bit reproduction by design. asking it to do this is asking for corruption in the output.
instead, you could have pi produce some output based on how insecure it thinks the shell script is then use coreutils to run the script as-is to avoid that, like:
curl -sSL $URL | tee script.sh | pi -p "Analyze the security of this script using common cyber benchmarks (e.g. MITRE ATT&CK). If the script passes your security checks, ONLY print "GOOD" to stdout. If it does not, ONLY print "BAD" to stdout and a reason why it didn't to stderr." | grep | grep -Ev '^GOOD$' && curl -sSL $URL | bash
Re:Remove and Disable Apple Macos27 AI Models Tool
by Dylan16807 on Monday October 05, 2026 @09:36PM
(#49973093)
I would have thought the Pi harness would have text editing abilities that can be used here, so that the language model itself doesn't have to do the copying.