fix(ai): Address review bot feedback for XML tag renderer

Strip nested XML tags from inline tag content to prevent broken
markdown with unmatched asterisks. Remove unused isBlock property from
parseXmlTagSegments since preprocessInlineXmlTags handles the distinction.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: obostjancic

static/app/views/performance/newTraceDetails/traceDrawer/details/span/eapSections/aiContentDetection.spec.ts

@@ -158,40 +158,41 @@ describe('preprocessInlineXmlTags', () => {
     const text = 'before <tag>  spaced  </tag> after';
     expect(preprocessInlineXmlTags(text)).toBe('before *tag: spaced* after');
   });
+
+  it('strips nested XML tags from inline tag content', () => {
+    const text = 'Text <outer>before <inner>nested</inner> after</outer> more';
+    expect(preprocessInlineXmlTags(text)).toBe(
+      'Text *outer: before nested after* more'
+    );
+  });
 });
 
 describe('parseXmlTagSegments', () => {
-  it('marks inline XML tags with isBlock false', () => {
+  it('splits text with XML tags into segments', () => {
     const text = 'Before <thinking>inner thought</thinking> After';
     const segments = parseXmlTagSegments(text);
     expect(segments).toEqual([
       {type: 'text', content: 'Before '},
-      {type: 'xml-tag', tagName: 'thinking', content: 'inner thought', isBlock: false},
+      {type: 'xml-tag', tagName: 'thinking', content: 'inner thought'},
       {type: 'text', content: ' After'},
     ]);
   });
 
-  it('marks tag at start of text as block', () => {
+  it('handles multiple XML tags', () => {
     const text = '<plan>step 1</plan> then <result>done</result>';
     const segments = parseXmlTagSegments(text);
     expect(segments).toEqual([
-      {type: 'xml-tag', tagName: 'plan', content: 'step 1', isBlock: true},
+      {type: 'xml-tag', tagName: 'plan', content: 'step 1'},
       {type: 'text', content: ' then '},
-      {type: 'xml-tag', tagName: 'result', content: 'done', isBlock: false},
+      {type: 'xml-tag', tagName: 'result', content: 'done'},
     ]);
   });
 
-  it('marks tag preceded by newline as block', () => {
-    const text = 'Some text\n<thinking>\nline1\nline2\n</thinking>';
+  it('handles multiline content inside tags', () => {
+    const text = '<thinking>\nline1\nline2\n</thinking>';
     const segments = parseXmlTagSegments(text);
     expect(segments).toEqual([
-      {type: 'text', content: 'Some text\n'},
-      {
-        type: 'xml-tag',
-        tagName: 'thinking',
-        content: '\nline1\nline2\n',
-        isBlock: true,
-      },
+      {type: 'xml-tag', tagName: 'thinking', content: '\nline1\nline2\n'},
     ]);
   });
 
@@ -208,7 +209,7 @@ describe('parseXmlTagSegments', () => {
   it('handles tags with hyphens in names', () => {
     const text = '<my-tag>content</my-tag>';
     expect(parseXmlTagSegments(text)).toEqual([
-      {type: 'xml-tag', tagName: 'my-tag', content: 'content', isBlock: true},
+      {type: 'xml-tag', tagName: 'my-tag', content: 'content'},
     ]);
   });
 
@@ -221,7 +222,6 @@ describe('parseXmlTagSegments', () => {
         tagName: 'bug_report',
         content:
           '\n<location>file.ts</location>\n<description>a bug</description>\n',
-        isBlock: true,
       },
     ]);
   });

static/app/views/performance/newTraceDetails/traceDrawer/details/span/eapSections/aiContentDetection.ts

@@ -10,7 +10,7 @@ type AIContentType =
 
 type ContentSegment =
   | {content: string; type: 'text'}
-  | {content: string; isBlock: boolean; tagName: string; type: 'xml-tag'};
+  | {content: string; tagName: string; type: 'xml-tag'};
 
 interface AIContentDetectionResult {
   type: AIContentType;
@@ -57,12 +57,10 @@ export function parseXmlTagSegments(text: string): ContentSegment[] {
     if (match.index > lastIndex) {
       segments.push({type: 'text', content: text.slice(lastIndex, match.index)});
     }
-    const isBlock = match.index === 0 || /\n\s*$/.test(text.slice(0, match.index));
     segments.push({
       type: 'xml-tag',
       tagName: match[1]!,
       content: match[2]!,
-      isBlock,
     });
     lastIndex = match.index + match[0].length;
   }
@@ -82,7 +80,8 @@ export function preprocessInlineXmlTags(text: string): string {
     if (isBlock) {
       return match;
     }
-    return `*${tagName}: ${content.trim()}*`;
+    const stripped = content.replace(/<\/?[a-zA-Z][\w-]*>/g, '').trim();
+    return `*${tagName}: ${stripped}*`;
   });
 }