SKILLs are the easiest way to configure agents, and you can get a lot of juice from a few well written SKILL files. Unfortunately, most SKILL files are not well written. They are way too long, they waste tokens explaining things that the model already knows, and many don't even follow the agentskills.io spec.
Linters help automate low hanging fruit changes that we find we constantly need to make.
You can run the linter as a one-liner: `npx nori-lint lint path/to/SKILL.md` or `npx nori-lint fix path/to/SKILL.md`
SKILL files are hard to lint deterministically, so there are several rules that require an LLM to do part of the review. You can set up a .nori-lint.json file to include an Anthropic API key, which will kick off a bunch of llm analysis rules.