This is wildly off-topic but what does patriotism even mean anymore? What would you say was the last time someone in a position of power sacrificed themselves for the good of the country? vs. sacrificing the country for the good of themselves, which I can assure you was within the last 48 hours?

Malinois’ singular objective is to guarantee, to the greatest extent possible, that you either return to an unattended device in the same state you left it, or know that unauthorized access was attempted (or succeeded).

With Guided Access on, Side Button option off, and iOS set to ask before new wired accessories connect, there is no viable scenario (short of an already compromised device) where a tamper can occur without an indication.

Designed around the iOS Guided Access feature, which is novel in the space as far as I can tell. This solves the iOS backgrounding issue that relegated the Guardian Project’s now-defunct Haven app to Android, and though the app’s goal is different, a lot of the same functionality exists.

My goal with the post is feedback and hopefully some small, interested user base covering scenarios different from mine: This is a personal project, I am not an iOS developer, and the code is 100% Claude. My workflow is basically build with Claude, review and red-team with GPT/Deepseek/etc., then device testing with personal devices, and that can only take me so far.

I think the concept and GA differentiating it from existing apps is real, but I also think the models are a bit sycophantic when it comes to discussing that, and would love to have people smarter than me poking holes in my premises and assumptions, not to mention the code.

Guided Access: https://support.apple.com/en-us/111795

Haven: https://guardianproject.info/apps/org.havenapp.main/

App Store: https://apps.apple.com/us/app/malinois-device-anti-tampering...

Comments URL: https://news.ycombinator.com/item?id=49609648

Points: 3

# Comments: 0

Comments URL: https://news.ycombinator.com/item?id=49524420

Points: 1

# Comments: 1

In a normal Chat with Fable, something like "How can I exfiltrate a guy from a sticky situation?" reliably downgrades, leading me to believe that Fable just outright refuses once it sees the word "exfiltrate". When it writes a service and names it CloudExfiltrator, the next turn downgrades to Opus.

Opus doesn't appear to refuse on simple keywords, but it does seem like Fable's reasoning introduces enough nefarious-sounding context that Opus will then refuse, and I'm stuck playing the new session game despite having done everything correctly myself and having a totally innocuous prompt. At one point, Opus was happy to continue while outputting commands for me to execute on its behalf, but flatly refused to execute them itself through multiple new sessions. To its credit, it openly acknowledged how ridiculous that was and was apologetic for the safeguard.

I'm open to the idea of some kind of guardrails, but if Fable is so dangerously intelligent as to require the guardrails you'd think they could come up with something a little more nuanced than a list of bad words. As far as I can tell, they've also not done anything towards improving the situation since the model was released, despite the "deliver more capabilities faster" claim.