When the same concept has two names within one context, both people and agents need to know both terms and remember that they refer to the same thing. Even if searching for both finds everything, it adds a memory and search burden. Someone who knows only one name may miss content under the other while believing the search is complete.
After unifying the domain language, I found that agents could locate relevant content more easily with grep and rg. Using a consistent name for the same concept in code and documentation reduces the need to guess and search for alternative names. Recording naming conventions in CONTEXT.md and applying them in code and documentation gives people and agents a shared vocabulary for understanding and searching.
This applies to concepts that are genuinely the same within one context. DDD's Ubiquitous Language also operates within a Bounded Context; it does not require merging concepts across different contexts. Consistent naming reduces omissions caused by inconsistent terminology, but it cannot by itself guarantee complete search results.
