nit

codex-rs/utils/stream-parser/README.md

@@ -90,15 +90,8 @@ assert_eq!(out.extracted.len(), 1);
 assert_eq!(out.extracted[0].content, "x");
 ```
 
-## Notes / limitations
+## Known limitations
 
 - Tags are matched literally and case-sensitively
-- No tag attributes
 - No nested tag support
-- Unterminated open tags are auto-closed on `finish()` (buffered content is returned as extracted)
-- `Utf8StreamParser::push_bytes(...)` rolls back the entire pushed chunk on invalid UTF-8 so the
-  wrapped parser does not observe a partial prefix from that chunk
-- `Utf8StreamParser::into_inner()` returns an error if undecoded UTF-8 bytes are still buffered;
-  use `into_inner_lossy()` only if you intentionally want to drop buffered partial bytes
-- `StreamTextParser::push_str(...)` accepts only valid UTF-8 (`&str`); use `Utf8StreamParser` if your
-  upstream source yields raw bytes
+- A stream can return empty objects.