Add directions around language handling for custom feeds

docs/starter-templates/custom-feeds.mdx

@@ -79,6 +79,10 @@ We strongly encourage that the cursor be _unique per feed item_ to prevent unexp
 We recommend, for instance, a compound cursor with a timestamp + a CID:
 `1683654690921::bafyreia3tbsfxe3cc75xrxyyn6qc42oupi73fxiox76prlyi5bpx7hr72u`
 
+#### Language handling
+
+Implemented correctly, requests to `getFeedSkeleton` should populate the `Accept-Language` header with comma-separated BCP-47 language codes e.g. `en,fr`. You should use this to return only posts that match those languages where possible. If language filtering is applied, you should in-turn populate the `Content-Language` header with the languages you used to filter.
+
 #### Suggestions and Examples
 
 How a feed generator fulfills the `getFeedSkeleton` request is completely at their discretion. At the simplest end, a Feed Generator could supply a "feed" that only contains some hardcoded posts.
@@ -89,4 +93,4 @@ Depending on your algorithm, you likely do not need to keep posts around for lon
 
 Some examples:
 - A community feed: Compile a list of DIDs within that community and filtering the firehose for all posts from users within that list.
-- A topical feed: Filter the algorithm for posts and pass the post text through some filtering mechanism (an LLM, a keyword matcher, etc.) that filters for the topic of your choice.
+- A topical feed: Filter the algorithm for posts and pass the post text through some filtering mechanism (an LLM, a keyword matcher, etc.) that filters for the topic of your choice.