Comment by rglover
1 month ago
For those who haven't gone down this rabbit hole like me yet: skills are just structured markdown files that describe how to handle a narrow-band task.
So, if I write my API endpoints a certain way, the skill would describe that specific process. Later, an agent can "see" this skill, load it when it's relevant to current chat context, and then do whatever is instructed.
Similar to "tool calls," but instead of being a function you can call, it's just instructions for how to perform that "skill."
At least for the agent I use (Cline), you can define skills either globally or locally (project level).
skills also have a header called "frontmatter" some piece of which us shared early in the context like a claude.md file
ive heard here that that skill loads can have a separate impact on the context like staying past a compact.
if you load a bunch of skills your session might end up with them permanently loaded.
i think they pair well with subagents, since the subagent can load the skill, and once its done with the work, can present just the results, and the orchestrator agent doesnt need to know about it