I don't exactly understand what you mean but it seems more like a problem with the auth token.

> Majorities of educators say student interactions drive their sense of mattering

Aren’t these two different things? While interactions with students are certainly important and represent the point where teachers can exert the greatest influence, the ultimate priority is how effectively the subject matter is conveyed. Planning and preparing lessons, for instance, may not provide a sense of mattering, but remains an essential prerequisite for being able to conduct the lesson in the first place and therefore does matter.

In other words, the danger I see here is that, for students, grades are ultimately what matter most. However, if a teacher has Claude create their teaching material and the students then solve the tasks via Claude, which are then corrected by Claude, then we have basically the same effect as a functional collapse in JEPA.

I doubt that steelworkers around 1920 typically went out dancing after their shifts. Back then, there was a lack of radios or sound systems, and live music was likely reserved for more upscale establishments. The average worker would more likely have gone to a beer hall or pub to have a drink with his mates.

I haven’t been able to quickly find any statistics for the U.S. that go back that far, but in Germany, around 67% of 18- to 25-year-olds were still consuming alcohol regularly in 1973, whereas that figure had dropped to about 28% by 2023. I’m not suggesting that this trend is inherently negative, but there is, at the very least, a correlation with a decline in social activities.

Like if I take his example: "Front-line staff may be skipping mandatory fields because the process adds fifteen minutes of friction to every customer interaction." First it's unclear whats the policy for those fields are when they are mandatory and also can be skipped. Then why are those mandatory if skipping them seems only lower friction with no other consequences? How should a model decide if it should enforce the policy for those fields, code an automation or just make them voluntary?

Next i'll do is to implement tavilly, exa, tinyfish etc. as search engines for searxng. No agents, no mcp, just their search api endpoint.

Sometimes I try other models but they always feel less concise and tend to not as strictly respect the instructions. Maybe the the top frontier models from openai / anthropic would not feel under this category but they are way to expensive on openrouter compared to my coding plan.

Same with mcp. I want them to use the jsdelivr cdn instead of them scraping github against the rate limit. etc. But if I dont explicitly state to strictly use the $%!$@@! mcp for searching in repositories they simply ignore the mcp and even if clearly instructed, they still often fall back to gh.

Putting every detailed instruction in the AGENTS.md would just unnecessarily bloat the context and it works well enough to just instruct them in the AGENTS.md when to use which skill. Yet I agree that Skills are not some voodoo magic to provide your model super capabilities.