Detect standalone confidence labels (HIGH, MEDIUM, LOW) in text nodes of the rendered HTML and wrap them in styled chip spans.
Matching is whole-word and case-insensitive. An Admiralty code in parentheses immediately following the label (e.g. "HIGH (A2)") is permitted but not required — the lookahead is optional, so any standalone occurrence is wrapped regardless of surrounding context.
Replacements are applied only to text nodes (not inside tag attributes) to avoid corrupting heading IDs, hrefs, or other attribute values.
Detect standalone confidence labels (HIGH, MEDIUM, LOW) in text nodes of the rendered HTML and wrap them in styled chip spans.
Matching is whole-word and case-insensitive. An Admiralty code in parentheses immediately following the label (e.g.
"HIGH (A2)") is permitted but not required — the lookahead is optional, so any standalone occurrence is wrapped regardless of surrounding context.Replacements are applied only to text nodes (not inside tag attributes) to avoid corrupting heading IDs, hrefs, or other attribute values.