diff --git a/web/js/utils.js b/web/js/utils.js
index 5207d54..50204ff 100644
--- a/web/js/utils.js
+++ b/web/js/utils.js
@@ -1,4 +1,6 @@
 // --- String Helper Functions ---
+const REG_WILDCARD_WEIGHTED_TAG = /(\d+)[_\s]*::(.*?)(?=\||$)/g;
+const REG_WILDCARD_SIMPLE_WORD = /[^{}_|]+/g;
 
 /**
  * Converts Hiragana to Katakana.
@@ -179,16 +181,113 @@ export function isValidTag(tag) {
     return true;
 }
 
+/**
+ * Recursively finds all words in a string, even those inside nested braces.
+ * This helps extract all possible tags from complex nested wildcards.
+ * @param {string} text The text to extract words from
+ * @param {number} baseStart The starting position of the text in the original string
+ * @returns {Array<{start: number, end: number, tag: string}>} Array of parsed tags
+ */
+function extractAllWords(text, baseStart) {
+    const result = [];
+    
+    // First, extract weighted tag patterns like "20::from above"
+    let weightMatch = REG_WILDCARD_WEIGHTED_TAG.exec(text);
+    while (weightMatch !== null) {
+        const tagText = weightMatch[2].trim();
+        
+        if (tagText) {
+            // Calculate position with original offsets
+            const fullMatchStart = baseStart + weightMatch.index;
+            const tagTextStart = fullMatchStart + weightMatch[0].indexOf(tagText);
+            const tagTextEnd = tagTextStart + tagText.length;
+            
+            result.push({
+                start: tagTextStart,
+                end: tagTextEnd,
+                tag: tagText
+            });
+        }
+
+        weightMatch = REG_WILDCARD_WEIGHTED_TAG.exec(text);
+    }
+    
+    // If no weighted tags were found, extract simple words
+    if (result.length === 0) {
+        // Regular expression to match words (sequences of non-whitespace characters)
+        // We consider a word to be any continuous sequence of characters that's not a space, pipe, or brace
+        const wordRegex = REG_WILDCARD_SIMPLE_WORD;
+        let match;
+        
+        // Find all standalone words in the text
+        while ((match = wordRegex.exec(text)) !== null) {
+            const wordStart = baseStart + match.index;
+            const wordEnd = wordStart + match[0].length;
+            
+            // Remove leading and trailing spaces from the matched word
+            const trimmedTag = match[0].trim();
+            const leadingSpaces = match[0].length - match[0].trimStart().length;
+            const adjustedStart = wordStart + leadingSpaces;
+            const adjustedEnd = wordStart + leadingSpaces + trimmedTag.length;
+            
+            result.push({
+                start: adjustedStart,
+                end: adjustedEnd,
+                tag: trimmedTag
+            });
+        }
+    }
+    
+    return result;
+}
+
+/**
+ * Parses a wildcard selection and returns individual tags.
+ * Supports syntax like {tag1|tag2|tag3} and {weight::tag1|weight::tag2}.
+ * Also handles nested wildcards like {tag1 {tag2|tag3}|tag4}.
+ * @param {string} tag The complete tag text that might contain a wildcard
+ * @param {number} startPos The starting position of the tag in the original text
+ * @param {number} endPos The ending position of the tag in the original text
+ * @returns {Array<{start: number, end: number, tag: string}>} Array of parsed tags or null
+ */
+function parseWildcardSelection(tag, startPos, endPos) {
+    // Trim the tag for matching but keep original position
+    const trimmedTag = tag.trim();
+    
+    // Check if this is a wildcard selection
+    if (!trimmedTag.startsWith('{') || !trimmedTag.endsWith('}')) {
+        return null; // Not a wildcard
+    }
+    
+    // Calculate position offsets for the trim operation
+    const leadingSpaces = tag.length - tag.trimStart().length;
+    const tagStart = startPos + leadingSpaces;
+    
+    // For nested wildcards, we'll extract all words from the content
+    // This treats each word as a separate tag, regardless of nesting
+    // Extract the content between the outermost braces
+    const wildcardContent = trimmedTag.substring(1, trimmedTag.length - 1);
+    
+    // Extract all words from the wildcard content, including those in nested structures
+    const allTags = extractAllWords(wildcardContent, tagStart + 1);
+    
+    return allTags.length > 0 ? allTags : null;
+}
+
 /**
  * Finds all tag positions in the given text.
  * Searches for tags separated by commas or newlines.
+ * Also handles wildcard selections in the format {tag1|tag2|tag3}.
  * @param {string} text The text to search in
  * @returns {Array<{start: number, end: number, tag: string}>} Array of tag positions and content
  */
 export function findAllTagPositions(text) {
+    if (!text) return [];
+    
     const positions = [];
     let startPos = 0;
-
+    
+    // Process text segment by segment (comma or newline separated)
     while (startPos < text.length) {
         // Skip any leading whitespace, commas, or newlines
         while (startPos < text.length &&
@@ -206,14 +305,26 @@ export function findAllTagPositions(text) {
         if (endPosNewline === -1) endPosNewline = text.length;
 
         const endPos = Math.min(endPosComma, endPosNewline);
-        const tag = text.substring(startPos, endPos);
+        const tagText = text.substring(startPos, endPos);
 
-        if (tag.trim().length > 0) {
-            positions.push({
-                start: startPos,
-                end: endPos,
-                tag: tag
-            });
+        if (tagText.trim().length > 0) {
+            const trimmedTag = tagText.trim();
+            
+            // Check if this is a wildcard selection
+            if (trimmedTag.startsWith('{') && trimmedTag.endsWith('}')) {
+                // Process wildcard using our existing wildcard parser
+                const wildcardTags = parseWildcardSelection(tagText, startPos, endPos);
+                if (wildcardTags) {
+                    positions.push(...wildcardTags);
+                }
+            } else {
+                // Normal tag, add it directly
+                positions.push({
+                    start: startPos,
+                    end: endPos,
+                    tag: tagText
+                });
+            }
         }
 
         // Move to the next tag
@@ -222,6 +333,7 @@ export function findAllTagPositions(text) {
 
     return positions;
 }
+
 /**
  * Extracts existing tags from the textarea with search normalization (possibly duplicated).
  * @param {HTMLTextAreaElement} textarea The textarea element to extract tags from