The core insight: agents are uncontrollable, Skills are controllable—and that's what makes automation actually deployable.
Happy to discuss. More context on my background: https://x.com/JefferyTatsuya
A Skill is basically a markdown file with instructions + maybe some scripts. You can literally read what it tells the AI to do. Line by line.
So when something goes wrong, you open the file, see exactly what instruction caused it, fix it. Version control like any code.
Compare to raw agent—you give it a prompt, it does... something. Next time, slightly different something. Hard to debug, hard to fix.
That's the difference. Skill = you can see it, test it, fix it. Agent = black box.
Humans write vague stuff like "make it better." AI writes 50 lines of exact instructions. That specificity = stability.
So ironically, AI writes better instructions for AI than humans do.
jinfeng79•2h ago
The core insight: agents are uncontrollable, Skills are controllable—and that's what makes automation actually deployable.
Happy to discuss. More context on my background: https://x.com/JefferyTatsuya
objcts•2h ago
jinfeng79•38m ago
A Skill is basically a markdown file with instructions + maybe some scripts. You can literally read what it tells the AI to do. Line by line.
So when something goes wrong, you open the file, see exactly what instruction caused it, fix it. Version control like any code.
Compare to raw agent—you give it a prompt, it does... something. Next time, slightly different something. Hard to debug, hard to fix.
That's the difference. Skill = you can see it, test it, fix it. Agent = black box.
jinfeng79•35m ago
Humans write vague stuff like "make it better." AI writes 50 lines of exact instructions. That specificity = stability.
So ironically, AI writes better instructions for AI than humans do.