Repository navigation
Are Agent Skills the right long-term abstraction for evolving enterprise systems? #659
hushoufeng
started this conversation in
General
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
We are building an internal support agent that needs to work across multiple business systems.
Today, different teams provide Skills, and a central agent decides which ones to use. Many Skills describe business flows, when to call specific APIs, and how to interpret the results.
This works, but as the number of Skills grows, I am becoming concerned about the long-term model:
This makes me wonder: Are Skills the right way to organize agent capabilities in the AI era, especially for enterprise systems that are constantly evolving?
Would it be more sustainable to let agents retrieve current information from the underlying systems, and reserve Skills for workflows or domain knowledge that genuinely improve results? Or is there something important about Skills that this view is missing?
I am not arguing that Skills are useless or that one approach is universally better. I would be interested in hearing from people who have built agents across multiple systems:
I would welcome both practical experiences and counterarguments.
All reactions