The human brain is the most complex thing in the known universe & using language is the most demanding thing we ever use it for. Whatever epistemology we choose, it's going to be extremely complex so there's never going to be a simple, straightforward answer to questions about what knowledge or skills are. Not in any useful way.
A common definition: Listening, speaking, reading, & writing are all examples of competencies. Competencies consist of combinations & configurations of knowledge, skills, & attitudes (KSAs). However, KSAs exist at different levels of analysis/complexity & exhibit dual structure, i.e. KSAs made out of KSAs, or KSAs nested within KSAs. It's complex!
Another common definition/distinction: Declarative knowledge is facts that you can talk about, describe, define, etc.. Procedural knowledge is knowing how to perform processes but not necessarily being able to talk about how you do them. Obviously, there isn't a clear dichotomy/dividing line between them. The declarative vs procedural distinction is sometimes useful for clarifying which learning & teaching strategies & techniques are more appropriate to a given learning objective. Again, it's complex!
A common claim in language learning is that words are knowledge & how we organise them into phrases, clauses, & sentences (syntax) is a mostly unconscious skill. This is a very narrow claim & cannot account for things like discourse, registers, & genres. Also, words exhibit syntactic features, just to muddy the waters further, e.g. "Tell me this." is correct but "Explain me this." isn't. Why? (Answer: See Adele Goldberg's work on construction grammar & usage-based models of language)
There's a variety of models of knowledge & language of which some are useful depending on what you're trying to do. An analogy is that models are like tools & you should have several tools in your tool box, e.g. hammer, screwdriver, & tape measure. Which one you choose depends on whether you're working with nails, screws, or lengths of wood.
A useful saying: "All models are wrong. Some are useful." - George Box (statistician)
I hope this helps! :)