A good 20 minute talk on how to empirically verify that the agent skills you create are actually doing what you think they're doing, and whether they're still even necessary:
1 Boost
0 Likes
And a reminder that we've got our own agent skill for navigating and querying the hundreds of tables and thousands of columns of data in PUDL, and all the rich metadata we've collected about them for years.
A related conversation with the founder of Tessl, on the professionalization of agent skills and the whole agentic software engineering stack. What does it take to test and maintain these things long term? What are all the layers?
https://podcasts.apple.com/us/podcast/you-cant-scale-vibes-guy-podjarny/id1574755368?i=1000786215912
The Analytics Engineering Podcast