Reading up on openskills.space

Agent skills (structured prompt instructions) are emerging as a key pattern for making LLMs reliable, but still lack standardization and face real-world system integration hurdles.