How to Implement Key-Value Storage Using `folly::ConcurrentSkipList`?

I'm trying to use folly::ConcurrentSkipList to implement a key-value store where each key maps to a single value, similar to std::map. However, the current implementation and documentation seem to focus on key-only usage (like std::set), and do not provide examples or guidance on how to store both key and value in the skip list.

From my understanding of folly::ConcurrentSkipList, it is designed for unique keys, like std::set, but I need to store both a key and its associated value. I am unsure how to modify the container or use an accessor to store and retrieve a key-value pair efficiently while ensuring the following:
- Lock-free or highly efficient thread-safe operations (insert, update, delete, and retrieval).
- Efficient handling of insertions, updates, deletions, and retrievals while maintaining the benefits of the skip list in terms of sorted access and concurrency.

Could you provide guidance on:
- How to implement key-value pairs using folly::ConcurrentSkipList?
- Best practices for handling operations (insert, update, delete, retrieve) while ensuring efficient thread-safe access?
- Ensuring the implementation remains lock-free or at least minimizes locking to ensure high concurrency and scalability?

No source file or test is named. Start by reading the folly::ConcurrentSkipList documentation and implementation to determine whether key-value storage is supported or requires a separate pattern. Done would be clear guidance covering representation and thread-safe insert, update, delete, and retrieval operations, ideally with an example.

Written by the indexing model from the issue text.