Interesting approach to the dynamic skill problem. I'm curious about the overhead. How much latency does the semantic search step usually add before the agent actually executes the task?
But why do agents need a skill to find skills? The security nightmare aside, don't they already have access or the ability to dig into what they need to do
For the security nightmare I have taken precautions, and by default only skills from trusted repos are given.
They could potentially websearch and find new skills, but that is tough, and of course will not be able to reach niche skills, that would be perfect for your use case. I have created semantic search with a ranking algorithm to give the appropriate skill in an instant.
I have indexed skills from trusted repos and github. And giving semantic search for the agent, along with ranking based on stars, forks, installs etc.
Trying to create a flywheel loop as well, where agents can rate skills and that gets incorporated in the ranking.
hm i see so a bit like pagerank for md files - would be curious about the equivalent to backlinks, what do you think that is for skills? Skills that embed other skills?
That is like a dependency thingy. Like browser harness by browser use is a skill. Using linkedin is a skill on top of that. Doing cold outreach on linkedin is probably a skill on top of that, and you can keep going.
I think if I install the cold outreach skill, I'll also have the browser harness skill as well, so the ranking algo will already rank it higher.
They could potentially websearch and find new skills, but that is tough, and of course will not be able to reach niche skills, that would be perfect for your use case. I have created semantic search with a ranking algorithm to give the appropriate skill in an instant.
I think if I install the cold outreach skill, I'll also have the browser harness skill as well, so the ranking algo will already rank it higher.