@j0hanz/superfetch 1.0.0 → 1.0.2

package/dist/transformers/jsonl.transformer.js

@@ -1,42 +1,47 @@
 import { config } from '../config/index.js';
 import { truncateText } from '../utils/sanitizer.js';
-import { logError } from '../services/logger.js';
 function truncateBlock(block) {
     const maxLength = config.extraction.maxBlockLength;
     switch (block.type) {
         case 'paragraph':
         case 'heading':
-        case 'code':
-            return { ...block, text: truncateText(block.text, maxLength) };
-        case 'list':
-            return {
-                ...block,
-                items: block.items.map((item) => truncateText(item, maxLength)),
-            };
+        case 'code': {
+            const truncated = truncateText(block.text, maxLength);
+            return truncated === block.text ? block : { ...block, text: truncated };
+        }
+        case 'list': {
+            const truncatedItems = block.items.map((item) => truncateText(item, maxLength));
+            const hasChanges = truncatedItems.some((item, i) => item !== block.items[i]);
+            return hasChanges ? { ...block, items: truncatedItems } : block;
+        }
         default:
             return block;
     }
 }
 export function toJsonl(blocks, metadata) {
     const lines = [];
-    if (metadata)
-        lines.push(JSON.stringify(metadata));
-    for (const block of blocks) {
-        lines.push(JSON.stringify(truncateBlock(block)));
+    // Minimal metadata - just title and URL for context
+    if (metadata) {
+        try {
+            const minimal = {
+                type: metadata.type,
+                title: metadata.title,
+                url: metadata.url,
+            };
+            lines.push(JSON.stringify(minimal));
+        }
+        catch {
+            // Skip invalid metadata
+        }
     }
-    return lines.join('\n');
-}
-export function fromJsonl(jsonl) {
-    const lines = jsonl.split('\n').filter((line) => line.trim());
-    const blocks = [];
-    for (const line of lines) {
+    for (const block of blocks) {
         try {
-            blocks.push(JSON.parse(line));
+            lines.push(JSON.stringify(truncateBlock(block)));
         }
-        catch (error) {
-            logError('Failed to parse JSONL line', error instanceof Error ? error : undefined);
+        catch {
+            // Skip blocks that fail to serialize
         }
     }
-    return blocks;
+    return lines.join('\n');
 }
 //# sourceMappingURL=jsonl.transformer.js.map

package/dist/transformers/jsonl.transformer.js.map

@@ -1 +1 @@
-{"version":3,"file":"jsonl.transformer.js","sourceRoot":"","sources":["../../src/transformers/jsonl.transformer.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAC5C,OAAO,EAAE,YAAY,EAAE,MAAM,uBAAuB,CAAC;AACrD,OAAO,EAAE,QAAQ,EAAE,MAAM,uBAAuB,CAAC;AAEjD,SAAS,aAAa,CAAC,KAAwB;IAC7C,MAAM,SAAS,GAAG,MAAM,CAAC,UAAU,CAAC,cAAc,CAAC;IAEnD,QAAQ,KAAK,CAAC,IAAI,EAAE,CAAC;QACnB,KAAK,WAAW,CAAC;QACjB,KAAK,SAAS,CAAC;QACf,KAAK,MAAM;YACT,OAAO,EAAE,GAAG,KAAK,EAAE,IAAI,EAAE,YAAY,CAAC,KAAK,CAAC,IAAI,EAAE,SAAS,CAAC,EAAE,CAAC;QACjE,KAAK,MAAM;YACT,OAAO;gBACL,GAAG,KAAK;gBACR,KAAK,EAAE,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,YAAY,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;aAChE,CAAC;QACJ;YACE,OAAO,KAAK,CAAC;IACjB,CAAC;AACH,CAAC;AAED,MAAM,UAAU,OAAO,CAAC,MAA2B,EAAE,QAAwB;IAC3E,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,IAAI,QAAQ;QAAE,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC,CAAC;IAEnD,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,aAAa,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;IACnD,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED,MAAM,UAAU,SAAS,CAAC,KAAa;IACrC,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,CAAC;IAC9D,MAAM,MAAM,GAAwB,EAAE,CAAC;IAEvC,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,IAAI,CAAC;YACH,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAsB,CAAC,CAAC;QACrD,CAAC;QAAC,OAAO,KAAK,EAAE,CAAC;YACf,QAAQ,CAAC,4BAA4B,EAAE,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC;QACrF,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC"}
+{"version":3,"file":"jsonl.transformer.js","sourceRoot":"","sources":["../../src/transformers/jsonl.transformer.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,MAAM,EAAE,MAAM,oBAAoB,CAAC;AAC5C,OAAO,EAAE,YAAY,EAAE,MAAM,uBAAuB,CAAC;AAErD,SAAS,aAAa,CAAC,KAAwB;IAC7C,MAAM,SAAS,GAAG,MAAM,CAAC,UAAU,CAAC,cAAc,CAAC;IAEnD,QAAQ,KAAK,CAAC,IAAI,EAAE,CAAC;QACnB,KAAK,WAAW,CAAC;QACjB,KAAK,SAAS,CAAC;QACf,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,SAAS,GAAG,YAAY,CAAC,KAAK,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;YACtD,OAAO,SAAS,KAAK,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,GAAG,KAAK,EAAE,IAAI,EAAE,SAAS,EAAE,CAAC;QAC1E,CAAC;QACD,KAAK,MAAM,CAAC,CAAC,CAAC;YACZ,MAAM,cAAc,GAAG,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,EAAE,CAC9C,YAAY,CAAC,IAAI,EAAE,SAAS,CAAC,CAC9B,CAAC;YACF,MAAM,UAAU,GAAG,cAAc,CAAC,IAAI,CACpC,CAAC,IAAI,EAAE,CAAC,EAAE,EAAE,CAAC,IAAI,KAAK,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CACrC,CAAC;YACF,OAAO,UAAU,CAAC,CAAC,CAAC,EAAE,GAAG,KAAK,EAAE,KAAK,EAAE,cAAc,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC;QAClE,CAAC;QACD;YACE,OAAO,KAAK,CAAC;IACjB,CAAC;AACH,CAAC;AAED,MAAM,UAAU,OAAO,CACrB,MAA2B,EAC3B,QAAwB;IAExB,MAAM,KAAK,GAAa,EAAE,CAAC;IAE3B,oDAAoD;IACpD,IAAI,QAAQ,EAAE,CAAC;QACb,IAAI,CAAC;YACH,MAAM,OAAO,GAAG;gBACd,IAAI,EAAE,QAAQ,CAAC,IAAI;gBACnB,KAAK,EAAE,QAAQ,CAAC,KAAK;gBACrB,GAAG,EAAE,QAAQ,CAAC,GAAG;aAClB,CAAC;YACF,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,OAAO,CAAC,CAAC,CAAC;QACtC,CAAC;QAAC,MAAM,CAAC;YACP,wBAAwB;QAC1B,CAAC;IACH,CAAC;IAED,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,IAAI,CAAC;YACH,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,aAAa,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;QACnD,CAAC;QAAC,MAAM,CAAC;YACP,qCAAqC;QACvC,CAAC;IACH,CAAC;IAED,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC"}

package/dist/transformers/markdown.transformer.d.ts

@@ -1,4 +1,3 @@
-import type { ContentBlockUnion, MetadataBlock } from '../types/index.js';
+import type { MetadataBlock } from '../types/index.js';
 export declare function htmlToMarkdown(html: string, metadata?: MetadataBlock): string;
-export declare function blocksToMarkdown(blocks: ContentBlockUnion[], metadata?: MetadataBlock): string;
 //# sourceMappingURL=markdown.transformer.d.ts.map

package/dist/transformers/markdown.transformer.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"markdown.transformer.d.ts","sourceRoot":"","sources":["../../src/transformers/markdown.transformer.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,iBAAiB,EAAE,aAAa,EAAE,MAAM,mBAAmB,CAAC;AA6F1E,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,CAAC,EAAE,aAAa,GAAG,MAAM,CAU7E;AAED,wBAAgB,gBAAgB,CAC9B,MAAM,EAAE,iBAAiB,EAAE,EAC3B,QAAQ,CAAC,EAAE,aAAa,GACvB,MAAM,CAcR"}
+{"version":3,"file":"markdown.transformer.d.ts","sourceRoot":"","sources":["../../src/transformers/markdown.transformer.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,aAAa,EAAE,MAAM,mBAAmB,CAAC;AAiDvD,wBAAgB,cAAc,CAAC,IAAI,EAAE,MAAM,EAAE,QAAQ,CAAC,EAAE,aAAa,GAAG,MAAM,CAkB7E"}

package/dist/transformers/markdown.transformer.js

@@ -3,20 +3,21 @@ const turndown = new TurndownService({
     headingStyle: 'atx',
     codeBlockStyle: 'fenced',
     emDelimiter: '_',
+    bulletListMarker: '-',
 });
-turndown.addRule('removeScripts', {
-    filter: ['script', 'style', 'noscript'],
+// Remove noise elements
+turndown.addRule('removeNoise', {
+    filter: ['script', 'style', 'noscript', 'nav', 'footer', 'aside', 'iframe'],
     replacement: () => '',
 });
-// Pre-compiled regex patterns for YAML value escaping (performance optimization)
+// Pre-compiled regex patterns
 const YAML_SPECIAL_CHARS = /[:[\]{}"\n\r'|>&*!?,#]/;
 const YAML_NUMERIC = /^[\d.]+$/;
 const YAML_RESERVED_WORDS = /^(true|false|null|yes|no|on|off)$/i;
 const ESCAPE_BACKSLASH = /\\/g;
 const ESCAPE_QUOTE = /"/g;
 const ESCAPE_NEWLINE = /\n/g;
-const ESCAPE_CARRIAGE = /\r/g;
-const ESCAPE_TAB = /\t/g;
+const MULTIPLE_NEWLINES = /\n{3,}/g;
 function escapeYamlValue(value) {
     const needsQuoting = YAML_SPECIAL_CHARS.test(value) ||
         value.startsWith(' ') ||
@@ -26,79 +27,35 @@ function escapeYamlValue(value) {
         YAML_RESERVED_WORDS.test(value);
     if (!needsQuoting)
         return value;
-    const escaped = value
+    return `"${value
         .replace(ESCAPE_BACKSLASH, '\\\\')
         .replace(ESCAPE_QUOTE, '\\"')
-        .replace(ESCAPE_NEWLINE, '\\n')
-        .replace(ESCAPE_CARRIAGE, '\\r')
-        .replace(ESCAPE_TAB, '\\t');
-    return `"${escaped}"`;
+        .replace(ESCAPE_NEWLINE, '\\n')}"`;
 }
 function createFrontmatter(metadata) {
     const lines = ['---'];
     if (metadata.title)
         lines.push(`title: ${escapeYamlValue(metadata.title)}`);
-    if (metadata.description)
-        lines.push(`description: ${escapeYamlValue(metadata.description)}`);
-    if (metadata.author)
-        lines.push(`author: ${escapeYamlValue(metadata.author)}`);
     if (metadata.url)
-        lines.push(`url: ${escapeYamlValue(metadata.url)}`);
-    if (metadata.fetchedAt)
-        lines.push(`fetched_at: ${escapeYamlValue(metadata.fetchedAt)}`);
+        lines.push(`source: ${escapeYamlValue(metadata.url)}`);
     lines.push('---');
     return lines.join('\n');
 }
-function tableToMarkdown(table) {
-    let markdown = '';
-    if (table.headers && table.headers.length > 0) {
-        markdown += '| ' + table.headers.join(' | ') + ' |\n';
-        markdown += '| ' + table.headers.map(() => '---').join(' | ') + ' |\n';
-    }
-    for (const row of table.rows) {
-        markdown += '| ' + row.join(' | ') + ' |\n';
+export function htmlToMarkdown(html, metadata) {
+    if (!html || typeof html !== 'string') {
+        return metadata ? createFrontmatter(metadata) + '\n\n' : '';
     }
-    return markdown.trim();
-}
-function blockToMarkdown(block) {
-    switch (block.type) {
-        case 'metadata':
-            return '';
-        case 'heading':
-            return '#'.repeat(block.level) + ' ' + block.text;
-        case 'paragraph':
-            return block.text;
-        case 'list':
-            return block.items
-                .map((item, index) => (block.ordered ? `${index + 1}. ` : '- ') + item)
-                .join('\n');
-        case 'code':
-            return '```' + (block.language || '') + '\n' + block.text + '\n```';
-        case 'table':
-            return tableToMarkdown(block);
-        case 'image':
-            return `![${block.alt || ''}](${block.src})`;
+    let content = '';
+    try {
+        content = turndown.turndown(html);
+        content = content.replace(MULTIPLE_NEWLINES, '\n\n').trim();
     }
-}
-export function htmlToMarkdown(html, metadata) {
-    let markdown = '';
-    if (metadata) {
-        markdown += createFrontmatter(metadata);
-        markdown += '\n\n';
+    catch {
+        return metadata ? createFrontmatter(metadata) + '\n\n' : '';
     }
-    markdown += turndown.turndown(html);
-    return markdown;
-}
-export function blocksToMarkdown(blocks, metadata) {
-    let markdown = '';
     if (metadata) {
-        markdown += createFrontmatter(metadata);
-        markdown += '\n\n';
-    }
-    for (const block of blocks) {
-        markdown += blockToMarkdown(block);
-        markdown += '\n\n';
+        return createFrontmatter(metadata) + '\n\n' + content;
     }
-    return markdown.trim();
+    return content;
 }
 //# sourceMappingURL=markdown.transformer.js.map

package/dist/transformers/markdown.transformer.js.map

@@ -1 +1 @@
-{"version":3,"file":"markdown.transformer.js","sourceRoot":"","sources":["../../src/transformers/markdown.transformer.ts"],"names":[],"mappings":"AAAA,OAAO,eAAe,MAAM,UAAU,CAAC;AAGvC,MAAM,QAAQ,GAAG,IAAI,eAAe,CAAC;IACnC,YAAY,EAAE,KAAK;IACnB,cAAc,EAAE,QAAQ;IACxB,WAAW,EAAE,GAAG;CACjB,CAAC,CAAC;AAEH,QAAQ,CAAC,OAAO,CAAC,eAAe,EAAE;IAChC,MAAM,EAAE,CAAC,QAAQ,EAAE,OAAO,EAAE,UAAU,CAAC;IACvC,WAAW,EAAE,GAAG,EAAE,CAAC,EAAE;CACtB,CAAC,CAAC;AAEH,iFAAiF;AACjF,MAAM,kBAAkB,GAAG,wBAAwB,CAAC;AACpD,MAAM,YAAY,GAAG,UAAU,CAAC;AAChC,MAAM,mBAAmB,GAAG,oCAAoC,CAAC;AACjE,MAAM,gBAAgB,GAAG,KAAK,CAAC;AAC/B,MAAM,YAAY,GAAG,IAAI,CAAC;AAC1B,MAAM,cAAc,GAAG,KAAK,CAAC;AAC7B,MAAM,eAAe,GAAG,KAAK,CAAC;AAC9B,MAAM,UAAU,GAAG,KAAK,CAAC;AAEzB,SAAS,eAAe,CAAC,KAAa;IACpC,MAAM,YAAY,GAChB,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC;QAC9B,KAAK,CAAC,UAAU,CAAC,GAAG,CAAC;QACrB,KAAK,CAAC,QAAQ,CAAC,GAAG,CAAC;QACnB,KAAK,KAAK,EAAE;QACZ,YAAY,CAAC,IAAI,CAAC,KAAK,CAAC;QACxB,mBAAmB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAElC,IAAI,CAAC,YAAY;QAAE,OAAO,KAAK,CAAC;IAEhC,MAAM,OAAO,GAAG,KAAK;SAClB,OAAO,CAAC,gBAAgB,EAAE,MAAM,CAAC;SACjC,OAAO,CAAC,YAAY,EAAE,KAAK,CAAC;SAC5B,OAAO,CAAC,cAAc,EAAE,KAAK,CAAC;SAC9B,OAAO,CAAC,eAAe,EAAE,KAAK,CAAC;SAC/B,OAAO,CAAC,UAAU,EAAE,KAAK,CAAC,CAAC;IAE9B,OAAO,IAAI,OAAO,GAAG,CAAC;AACxB,CAAC;AAED,SAAS,iBAAiB,CAAC,QAAuB;IAChD,MAAM,KAAK,GAAG,CAAC,KAAK,CAAC,CAAC;IAEtB,IAAI,QAAQ,CAAC,KAAK;QAAE,KAAK,CAAC,IAAI,CAAC,UAAU,eAAe,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;IAC5E,IAAI,QAAQ,CAAC,WAAW;QAAE,KAAK,CAAC,IAAI,CAAC,gBAAgB,eAAe,CAAC,QAAQ,CAAC,WAAW,CAAC,EAAE,CAAC,CAAC;IAC9F,IAAI,QAAQ,CAAC,MAAM;QAAE,KAAK,CAAC,IAAI,CAAC,WAAW,eAAe,CAAC,QAAQ,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC;IAC/E,IAAI,QAAQ,CAAC,GAAG;QAAE,KAAK,CAAC,IAAI,CAAC,QAAQ,eAAe,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;IACtE,IAAI,QAAQ,CAAC,SAAS;QAAE,KAAK,CAAC,IAAI,CAAC,eAAe,eAAe,CAAC,QAAQ,CAAC,SAAS,CAAC,EAAE,CAAC,CAAC;IAEzF,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAClB,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED,SAAS,eAAe,CAAC,KAA+C;IACtE,IAAI,QAAQ,GAAG,EAAE,CAAC;IAElB,IAAI,KAAK,CAAC,OAAO,IAAI,KAAK,CAAC,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QAC9C,QAAQ,IAAI,IAAI,GAAG,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,MAAM,CAAC;QACtD,QAAQ,IAAI,IAAI,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,MAAM,CAAC;IACzE,CAAC;IAED,KAAK,MAAM,GAAG,IAAI,KAAK,CAAC,IAAI,EAAE,CAAC;QAC7B,QAAQ,IAAI,IAAI,GAAG,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,MAAM,CAAC;IAC9C,CAAC;IAED,OAAO,QAAQ,CAAC,IAAI,EAAE,CAAC;AACzB,CAAC;AAED,SAAS,eAAe,CAAC,KAAwB;IAC/C,QAAQ,KAAK,CAAC,IAAI,EAAE,CAAC;QACnB,KAAK,UAAU;YACb,OAAO,EAAE,CAAC;QACZ,KAAK,SAAS;YACZ,OAAO,GAAG,CAAC,MAAM,CAAC,KAAK,CAAC,KAAK,CAAC,GAAG,GAAG,GAAG,KAAK,CAAC,IAAI,CAAC;QACpD,KAAK,WAAW;YACd,OAAO,KAAK,CAAC,IAAI,CAAC;QACpB,KAAK,MAAM;YACT,OAAO,KAAK,CAAC,KAAK;iBACf,GAAG,CAAC,CAAC,IAAI,EAAE,KAAK,EAAE,EAAE,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,GAAG,KAAK,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,IAAI,CAAC;iBACtE,IAAI,CAAC,IAAI,CAAC,CAAC;QAChB,KAAK,MAAM;YACT,OAAO,KAAK,GAAG,CAAC,KAAK,CAAC,QAAQ,IAAI,EAAE,CAAC,GAAG,IAAI,GAAG,KAAK,CAAC,IAAI,GAAG,OAAO,CAAC;QACtE,KAAK,OAAO;YACV,OAAO,eAAe,CAAC,KAAK,CAAC,CAAC;QAChC,KAAK,OAAO;YACV,OAAO,KAAK,KAAK,CAAC,GAAG,IAAI,EAAE,KAAK,KAAK,CAAC,GAAG,GAAG,CAAC;IACjD,CAAC;AACH,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,IAAY,EAAE,QAAwB;IACnE,IAAI,QAAQ,GAAG,EAAE,CAAC;IAElB,IAAI,QAAQ,EAAE,CAAC;QACb,QAAQ,IAAI,iBAAiB,CAAC,QAAQ,CAAC,CAAC;QACxC,QAAQ,IAAI,MAAM,CAAC;IACrB,CAAC;IAED,QAAQ,IAAI,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;IACpC,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,MAAM,UAAU,gBAAgB,CAC9B,MAA2B,EAC3B,QAAwB;IAExB,IAAI,QAAQ,GAAG,EAAE,CAAC;IAElB,IAAI,QAAQ,EAAE,CAAC;QACb,QAAQ,IAAI,iBAAiB,CAAC,QAAQ,CAAC,CAAC;QACxC,QAAQ,IAAI,MAAM,CAAC;IACrB,CAAC;IAED,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,QAAQ,IAAI,eAAe,CAAC,KAAK,CAAC,CAAC;QACnC,QAAQ,IAAI,MAAM,CAAC;IACrB,CAAC;IAED,OAAO,QAAQ,CAAC,IAAI,EAAE,CAAC;AACzB,CAAC"}
+{"version":3,"file":"markdown.transformer.js","sourceRoot":"","sources":["../../src/transformers/markdown.transformer.ts"],"names":[],"mappings":"AAAA,OAAO,eAAe,MAAM,UAAU,CAAC;AAGvC,MAAM,QAAQ,GAAG,IAAI,eAAe,CAAC;IACnC,YAAY,EAAE,KAAK;IACnB,cAAc,EAAE,QAAQ;IACxB,WAAW,EAAE,GAAG;IAChB,gBAAgB,EAAE,GAAG;CACtB,CAAC,CAAC;AAEH,wBAAwB;AACxB,QAAQ,CAAC,OAAO,CAAC,aAAa,EAAE;IAC9B,MAAM,EAAE,CAAC,QAAQ,EAAE,OAAO,EAAE,UAAU,EAAE,KAAK,EAAE,QAAQ,EAAE,OAAO,EAAE,QAAQ,CAAC;IAC3E,WAAW,EAAE,GAAG,EAAE,CAAC,EAAE;CACtB,CAAC,CAAC;AAEH,8BAA8B;AAC9B,MAAM,kBAAkB,GAAG,wBAAwB,CAAC;AACpD,MAAM,YAAY,GAAG,UAAU,CAAC;AAChC,MAAM,mBAAmB,GAAG,oCAAoC,CAAC;AACjE,MAAM,gBAAgB,GAAG,KAAK,CAAC;AAC/B,MAAM,YAAY,GAAG,IAAI,CAAC;AAC1B,MAAM,cAAc,GAAG,KAAK,CAAC;AAC7B,MAAM,iBAAiB,GAAG,SAAS,CAAC;AAEpC,SAAS,eAAe,CAAC,KAAa;IACpC,MAAM,YAAY,GAChB,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC;QAC9B,KAAK,CAAC,UAAU,CAAC,GAAG,CAAC;QACrB,KAAK,CAAC,QAAQ,CAAC,GAAG,CAAC;QACnB,KAAK,KAAK,EAAE;QACZ,YAAY,CAAC,IAAI,CAAC,KAAK,CAAC;QACxB,mBAAmB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAElC,IAAI,CAAC,YAAY;QAAE,OAAO,KAAK,CAAC;IAEhC,OAAO,IAAI,KAAK;SACb,OAAO,CAAC,gBAAgB,EAAE,MAAM,CAAC;SACjC,OAAO,CAAC,YAAY,EAAE,KAAK,CAAC;SAC5B,OAAO,CAAC,cAAc,EAAE,KAAK,CAAC,GAAG,CAAC;AACvC,CAAC;AAED,SAAS,iBAAiB,CAAC,QAAuB;IAChD,MAAM,KAAK,GAAG,CAAC,KAAK,CAAC,CAAC;IACtB,IAAI,QAAQ,CAAC,KAAK;QAAE,KAAK,CAAC,IAAI,CAAC,UAAU,eAAe,CAAC,QAAQ,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC;IAC5E,IAAI,QAAQ,CAAC,GAAG;QAAE,KAAK,CAAC,IAAI,CAAC,WAAW,eAAe,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;IACzE,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAClB,OAAO,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAC1B,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,IAAY,EAAE,QAAwB;IACnE,IAAI,CAAC,IAAI,IAAI,OAAO,IAAI,KAAK,QAAQ,EAAE,CAAC;QACtC,OAAO,QAAQ,CAAC,CAAC,CAAC,iBAAiB,CAAC,QAAQ,CAAC,GAAG,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;IAC9D,CAAC;IAED,IAAI,OAAO,GAAG,EAAE,CAAC;IACjB,IAAI,CAAC;QACH,OAAO,GAAG,QAAQ,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC;QAClC,OAAO,GAAG,OAAO,CAAC,OAAO,CAAC,iBAAiB,EAAE,MAAM,CAAC,CAAC,IAAI,EAAE,CAAC;IAC9D,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,QAAQ,CAAC,CAAC,CAAC,iBAAiB,CAAC,QAAQ,CAAC,GAAG,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC;IAC9D,CAAC;IAED,IAAI,QAAQ,EAAE,CAAC;QACb,OAAO,iBAAiB,CAAC,QAAQ,CAAC,GAAG,MAAM,GAAG,OAAO,CAAC;IACxD,CAAC;IAED,OAAO,OAAO,CAAC;AACjB,CAAC"}

package/dist/types/content.types.d.ts

@@ -1,4 +1,4 @@
-export type ContentBlockType = 'metadata' | 'heading' | 'paragraph' | 'list' | 'code' | 'table' | 'image';
+type ContentBlockType = 'metadata' | 'heading' | 'paragraph' | 'list' | 'code' | 'table' | 'image';
 interface ContentBlock {
     type: ContentBlockType;
 }

package/dist/types/content.types.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"content.types.d.ts","sourceRoot":"","sources":["../../src/types/content.types.ts"],"names":[],"mappings":"AACA,MAAM,MAAM,gBAAgB,GACxB,UAAU,GACV,SAAS,GACT,WAAW,GACX,MAAM,GACN,MAAM,GACN,OAAO,GACP,OAAO,CAAC;AAGZ,UAAU,YAAY;IACpB,IAAI,EAAE,gBAAgB,CAAC;CACxB;AAGD,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,IAAI,EAAE,UAAU,CAAC;IACjB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,GAAG,EAAE,MAAM,CAAC;IACZ,SAAS,EAAE,MAAM,CAAC;CACnB;AAGD,MAAM,WAAW,YAAa,SAAQ,YAAY;IAChD,IAAI,EAAE,SAAS,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,WAAW,cAAe,SAAQ,YAAY;IAClD,IAAI,EAAE,WAAW,CAAC;IAClB,IAAI,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,OAAO,CAAC;IACjB,KAAK,EAAE,MAAM,EAAE,CAAC;CACjB;AAGD,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,IAAI,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,WAAW,UAAW,SAAQ,YAAY;IAC9C,IAAI,EAAE,OAAO,CAAC;IACd,OAAO,CAAC,EAAE,MAAM,EAAE,CAAC;IACnB,IAAI,EAAE,MAAM,EAAE,EAAE,CAAC;CAClB;AAGD,MAAM,WAAW,UAAW,SAAQ,YAAY;IAC9C,IAAI,EAAE,OAAO,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;IACZ,GAAG,CAAC,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,MAAM,iBAAiB,GACzB,aAAa,GACb,YAAY,GACZ,cAAc,GACd,SAAS,GACT,SAAS,GACT,UAAU,GACV,UAAU,CAAC;AAGf,MAAM,WAAW,gBAAgB;IAC/B,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,MAAM,CAAC;IAChB,WAAW,EAAE,MAAM,CAAC;IACpB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;CACnB;AAGD,MAAM,WAAW,UAAU;IACzB,GAAG,EAAE,MAAM,CAAC;IACZ,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,MAAM,CAAC;CACnB;AAGD,MAAM,WAAW,aAAa;IAC5B,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,UAAU,GAAG,UAAU,CAAC;CAC/B"}
+{"version":3,"file":"content.types.d.ts","sourceRoot":"","sources":["../../src/types/content.types.ts"],"names":[],"mappings":"AACA,KAAK,gBAAgB,GACjB,UAAU,GACV,SAAS,GACT,WAAW,GACX,MAAM,GACN,MAAM,GACN,OAAO,GACP,OAAO,CAAC;AAGZ,UAAU,YAAY;IACpB,IAAI,EAAE,gBAAgB,CAAC;CACxB;AAGD,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,IAAI,EAAE,UAAU,CAAC;IACjB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,GAAG,EAAE,MAAM,CAAC;IACZ,SAAS,EAAE,MAAM,CAAC;CACnB;AAGD,MAAM,WAAW,YAAa,SAAQ,YAAY;IAChD,IAAI,EAAE,SAAS,CAAC;IAChB,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,WAAW,cAAe,SAAQ,YAAY;IAClD,IAAI,EAAE,WAAW,CAAC;IAClB,IAAI,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,OAAO,CAAC;IACjB,KAAK,EAAE,MAAM,EAAE,CAAC;CACjB;AAGD,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,IAAI,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,WAAW,UAAW,SAAQ,YAAY;IAC9C,IAAI,EAAE,OAAO,CAAC;IACd,OAAO,CAAC,EAAE,MAAM,EAAE,CAAC;IACnB,IAAI,EAAE,MAAM,EAAE,EAAE,CAAC;CAClB;AAGD,MAAM,WAAW,UAAW,SAAQ,YAAY;IAC9C,IAAI,EAAE,OAAO,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;IACZ,GAAG,CAAC,EAAE,MAAM,CAAC;CACd;AAGD,MAAM,MAAM,iBAAiB,GACzB,aAAa,GACb,YAAY,GACZ,cAAc,GACd,SAAS,GACT,SAAS,GACT,UAAU,GACV,UAAU,CAAC;AAGf,MAAM,WAAW,gBAAgB;IAC/B,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,OAAO,EAAE,MAAM,CAAC;IAChB,WAAW,EAAE,MAAM,CAAC;IACpB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;CACnB;AAGD,MAAM,WAAW,UAAU;IACzB,GAAG,EAAE,MAAM,CAAC;IACZ,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,MAAM,CAAC;IAClB,SAAS,EAAE,MAAM,CAAC;CACnB;AAGD,MAAM,WAAW,aAAa;IAC5B,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,UAAU,GAAG,UAAU,CAAC;CAC/B"}

package/dist/utils/sanitizer.d.ts

@@ -1,9 +1,13 @@
 /**
- * Sanitizes text content by removing extra whitespace and special characters
+ * Sanitizes text content by collapsing whitespace and trimming
+ * Returns empty string for null/undefined input
  */
-export declare function sanitizeText(text: string): string;
+export declare function sanitizeText(text: string | null | undefined): string;
 /**
  * Truncates text to a maximum length with ellipsis
+ * @param text - Text to truncate
+ * @param maxLength - Maximum length (must be > 3 to accommodate ellipsis)
+ * @returns Truncated text with ellipsis if needed
  */
 export declare function truncateText(text: string, maxLength: number): string;
 //# sourceMappingURL=sanitizer.d.ts.map

package/dist/utils/sanitizer.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"sanitizer.d.ts","sourceRoot":"","sources":["../../src/utils/sanitizer.ts"],"names":[],"mappings":"AAAA;;GAEG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAKjD;AAED;;GAEG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,GAAG,MAAM,CAKpE"}
+{"version":3,"file":"sanitizer.d.ts","sourceRoot":"","sources":["../../src/utils/sanitizer.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG,IAAI,GAAG,SAAS,GAAG,MAAM,CAIpE;AAED;;;;;GAKG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,GAAG,MAAM,CAQpE"}

package/dist/utils/sanitizer.js

@@ -1,16 +1,24 @@
 /**
- * Sanitizes text content by removing extra whitespace and special characters
+ * Sanitizes text content by collapsing whitespace and trimming
+ * Returns empty string for null/undefined input
  */
 export function sanitizeText(text) {
-    return text
-        .replace(/\s+/g, ' ') // Replace multiple spaces with single space
-        .replace(/\n\s*\n/g, '\n') // Remove empty lines
-        .trim();
+    if (text == null)
+        return '';
+    if (typeof text !== 'string')
+        return String(text);
+    return text.replace(/\s+/g, ' ').trim();
 }
 /**
  * Truncates text to a maximum length with ellipsis
+ * @param text - Text to truncate
+ * @param maxLength - Maximum length (must be > 3 to accommodate ellipsis)
+ * @returns Truncated text with ellipsis if needed
  */
 export function truncateText(text, maxLength) {
+    if (maxLength < 4) {
+        return text.length > 0 ? text.charAt(0) : '';
+    }
     if (text.length <= maxLength) {
         return text;
     }

package/dist/utils/sanitizer.js.map

@@ -1 +1 @@
-{"version":3,"file":"sanitizer.js","sourceRoot":"","sources":["../../src/utils/sanitizer.ts"],"names":[],"mappings":"AAAA;;GAEG;AACH,MAAM,UAAU,YAAY,CAAC,IAAY;IACvC,OAAO,IAAI;SACR,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,4CAA4C;SACjE,OAAO,CAAC,UAAU,EAAE,IAAI,CAAC,CAAC,qBAAqB;SAC/C,IAAI,EAAE,CAAC;AACZ,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,YAAY,CAAC,IAAY,EAAE,SAAiB;IAC1D,IAAI,IAAI,CAAC,MAAM,IAAI,SAAS,EAAE,CAAC;QAC7B,OAAO,IAAI,CAAC;IACd,CAAC;IACD,OAAO,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,SAAS,GAAG,CAAC,CAAC,GAAG,KAAK,CAAC;AAClD,CAAC"}
+{"version":3,"file":"sanitizer.js","sourceRoot":"","sources":["../../src/utils/sanitizer.ts"],"names":[],"mappings":"AAAA;;;GAGG;AACH,MAAM,UAAU,YAAY,CAAC,IAA+B;IAC1D,IAAI,IAAI,IAAI,IAAI;QAAE,OAAO,EAAE,CAAC;IAC5B,IAAI,OAAO,IAAI,KAAK,QAAQ;QAAE,OAAO,MAAM,CAAC,IAAI,CAAC,CAAC;IAClD,OAAO,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;AAC1C,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,YAAY,CAAC,IAAY,EAAE,SAAiB;IAC1D,IAAI,SAAS,GAAG,CAAC,EAAE,CAAC;QAClB,OAAO,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;IAC/C,CAAC;IACD,IAAI,IAAI,CAAC,MAAM,IAAI,SAAS,EAAE,CAAC;QAC7B,OAAO,IAAI,CAAC;IACd,CAAC;IACD,OAAO,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,SAAS,GAAG,CAAC,CAAC,GAAG,KAAK,CAAC;AAClD,CAAC"}

package/dist/utils/tool-error-handler.d.ts

@@ -0,0 +1,18 @@
+/** MCP SDK-compatible error response (index signature required by SDK) */
+export type ToolErrorResponse = {
+    [x: string]: unknown;
+    content: {
+        type: 'text';
+        text: string;
+    }[];
+    structuredContent: {
+        [x: string]: unknown;
+        error: string;
+        url: string;
+        errorCode: string;
+    };
+    isError: true;
+};
+export declare function createToolErrorResponse(message: string, url: string, code: string): ToolErrorResponse;
+export declare function handleToolError(error: unknown, url: string, fallbackMessage?: string): ToolErrorResponse;
+//# sourceMappingURL=tool-error-handler.d.ts.map

package/dist/utils/tool-error-handler.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"tool-error-handler.d.ts","sourceRoot":"","sources":["../../src/utils/tool-error-handler.ts"],"names":[],"mappings":"AAOA,0EAA0E;AAC1E,MAAM,MAAM,iBAAiB,GAAG;IAC9B,CAAC,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC;IACrB,OAAO,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC1C,iBAAiB,EAAE;QACjB,CAAC,CAAC,EAAE,MAAM,GAAG,OAAO,CAAC;QACrB,KAAK,EAAE,MAAM,CAAC;QACd,GAAG,EAAE,MAAM,CAAC;QACZ,SAAS,EAAE,MAAM,CAAC;KACnB,CAAC;IACF,OAAO,EAAE,IAAI,CAAC;CACf,CAAC;AAEF,wBAAgB,uBAAuB,CACrC,OAAO,EAAE,MAAM,EACf,GAAG,EAAE,MAAM,EACX,IAAI,EAAE,MAAM,GACX,iBAAiB,CAOnB;AAED,wBAAgB,eAAe,CAC7B,KAAK,EAAE,OAAO,EACd,GAAG,EAAE,MAAM,EACX,eAAe,SAAqB,GACnC,iBAAiB,CAqBnB"}

package/dist/utils/tool-error-handler.js

@@ -0,0 +1,27 @@
+import { AppError, UrlValidationError, FetchError, TimeoutError, } from '../errors/index.js';
+export function createToolErrorResponse(message, url, code) {
+    const structuredContent = { error: message, url, errorCode: code };
+    return {
+        content: [{ type: 'text', text: JSON.stringify(structuredContent) }],
+        structuredContent,
+        isError: true,
+    };
+}
+export function handleToolError(error, url, fallbackMessage = 'Operation failed') {
+    if (error instanceof UrlValidationError) {
+        return createToolErrorResponse(error.message, url, 'INVALID_URL');
+    }
+    if (error instanceof TimeoutError) {
+        return createToolErrorResponse(`Request timed out after ${error.timeoutMs}ms`, url, 'TIMEOUT');
+    }
+    if (error instanceof FetchError) {
+        const code = error.httpStatus ? `HTTP_${error.httpStatus}` : 'FETCH_ERROR';
+        return createToolErrorResponse(error.message, url, code);
+    }
+    if (error instanceof AppError) {
+        return createToolErrorResponse(error.message, url, error.code);
+    }
+    const message = error instanceof Error ? error.message : 'Unknown error';
+    return createToolErrorResponse(`${fallbackMessage}: ${message}`, url, 'UNKNOWN_ERROR');
+}
+//# sourceMappingURL=tool-error-handler.js.map

package/dist/utils/tool-error-handler.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"tool-error-handler.js","sourceRoot":"","sources":["../../src/utils/tool-error-handler.ts"],"names":[],"mappings":"AAAA,OAAO,EACL,QAAQ,EACR,kBAAkB,EAClB,UAAU,EACV,YAAY,GACb,MAAM,oBAAoB,CAAC;AAe5B,MAAM,UAAU,uBAAuB,CACrC,OAAe,EACf,GAAW,EACX,IAAY;IAEZ,MAAM,iBAAiB,GAAG,EAAE,KAAK,EAAE,OAAO,EAAE,GAAG,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC;IACnE,OAAO;QACL,OAAO,EAAE,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC,iBAAiB,CAAC,EAAE,CAAC;QACpE,iBAAiB;QACjB,OAAO,EAAE,IAAI;KACd,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,eAAe,CAC7B,KAAc,EACd,GAAW,EACX,eAAe,GAAG,kBAAkB;IAEpC,IAAI,KAAK,YAAY,kBAAkB,EAAE,CAAC;QACxC,OAAO,uBAAuB,CAAC,KAAK,CAAC,OAAO,EAAE,GAAG,EAAE,aAAa,CAAC,CAAC;IACpE,CAAC;IACD,IAAI,KAAK,YAAY,YAAY,EAAE,CAAC;QAClC,OAAO,uBAAuB,CAC5B,2BAA2B,KAAK,CAAC,SAAS,IAAI,EAC9C,GAAG,EACH,SAAS,CACV,CAAC;IACJ,CAAC;IACD,IAAI,KAAK,YAAY,UAAU,EAAE,CAAC;QAChC,MAAM,IAAI,GAAG,KAAK,CAAC,UAAU,CAAC,CAAC,CAAC,QAAQ,KAAK,CAAC,UAAU,EAAE,CAAC,CAAC,CAAC,aAAa,CAAC;QAC3E,OAAO,uBAAuB,CAAC,KAAK,CAAC,OAAO,EAAE,GAAG,EAAE,IAAI,CAAC,CAAC;IAC3D,CAAC;IACD,IAAI,KAAK,YAAY,QAAQ,EAAE,CAAC;QAC9B,OAAO,uBAAuB,CAAC,KAAK,CAAC,OAAO,EAAE,GAAG,EAAE,KAAK,CAAC,IAAI,CAAC,CAAC;IACjE,CAAC;IAED,MAAM,OAAO,GAAG,KAAK,YAAY,KAAK,CAAC,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,eAAe,CAAC;IACzE,OAAO,uBAAuB,CAAC,GAAG,eAAe,KAAK,OAAO,EAAE,EAAE,GAAG,EAAE,eAAe,CAAC,CAAC;AACzF,CAAC"}

package/dist/utils/url-validator.d.ts

@@ -1,5 +1,6 @@
 /**
  * Validates and normalizes a URL, blocking SSRF attack vectors
+ * @throws {ValidationError} if URL is empty or too long
  * @throws {UrlValidationError} if URL is invalid or blocked
  */
 export declare function validateAndNormalizeUrl(urlString: string): string;

package/dist/utils/url-validator.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"url-validator.d.ts","sourceRoot":"","sources":["../../src/utils/url-validator.ts"],"names":[],"mappings":"AAgCA;;;GAGG;AACH,wBAAgB,uBAAuB,CAAC,SAAS,EAAE,MAAM,GAAG,MAAM,CAoCjE;AAED;;GAEG;AACH,wBAAgB,aAAa,CAAC,GAAG,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,OAAO,CAQnE"}
+{"version":3,"file":"url-validator.d.ts","sourceRoot":"","sources":["../../src/utils/url-validator.ts"],"names":[],"mappings":"AAyCA;;;;GAIG;AACH,wBAAgB,uBAAuB,CAAC,SAAS,EAAE,MAAM,GAAG,MAAM,CA2EjE;AAED;;GAEG;AACH,wBAAgB,aAAa,CAAC,GAAG,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,GAAG,OAAO,CAQnE"}

package/dist/utils/url-validator.js

@@ -1,4 +1,6 @@
-import { UrlValidationError } from '../errors/app-error.js';
+import { UrlValidationError, ValidationError } from '../errors/app-error.js';
+// Maximum URL length to prevent DoS attacks
+const MAX_URL_LENGTH = 2048;
 // Blocked hosts to prevent SSRF attacks
 const BLOCKED_HOSTS = new Set([
     'localhost',
@@ -8,6 +10,8 @@ const BLOCKED_HOSTS = new Set([
     '169.254.169.254', // AWS metadata endpoint
     'metadata.google.internal', // GCP metadata
     'metadata.azure.com', // Azure metadata
+    '100.100.100.200', // Alibaba Cloud metadata
+    'instance-data', // Common cloud metadata hostname
 ]);
 // Blocked IP patterns (private networks)
 const BLOCKED_IP_PATTERNS = [
@@ -19,6 +23,10 @@ const BLOCKED_IP_PATTERNS = [
     /^169\.254\./, // Link-local
     /^fc00:/i, // IPv6 unique local
     /^fe80:/i, // IPv6 link-local
+    /^::ffff:127\./, // IPv4-mapped IPv6 loopback
+    /^::ffff:10\./, // IPv4-mapped IPv6 private
+    /^::ffff:172\.(1[6-9]|2\d|3[01])\./, // IPv4-mapped IPv6 private
+    /^::ffff:192\.168\./, // IPv4-mapped IPv6 private
 ];
 /**
  * Checks if a hostname matches blocked IP patterns
@@ -28,28 +36,53 @@ function isBlockedIp(hostname) {
 }
 /**
  * Validates and normalizes a URL, blocking SSRF attack vectors
+ * @throws {ValidationError} if URL is empty or too long
  * @throws {UrlValidationError} if URL is invalid or blocked
  */
 export function validateAndNormalizeUrl(urlString) {
+    // Check for empty or whitespace-only input
+    if (!urlString || typeof urlString !== 'string') {
+        throw new ValidationError('URL is required');
+    }
+    const trimmedUrl = urlString.trim();
+    if (!trimmedUrl) {
+        throw new ValidationError('URL cannot be empty');
+    }
+    // Check URL length to prevent DoS
+    if (trimmedUrl.length > MAX_URL_LENGTH) {
+        throw new ValidationError(`URL exceeds maximum length of ${MAX_URL_LENGTH} characters`, { length: trimmedUrl.length, maxLength: MAX_URL_LENGTH });
+    }
     let url;
     try {
-        url = new URL(urlString);
+        url = new URL(trimmedUrl);
     }
     catch {
-        throw new UrlValidationError(`Invalid URL format`, urlString);
+        throw new UrlValidationError(`Invalid URL format`, trimmedUrl);
     }
     // Only allow HTTP(S) protocols
     if (url.protocol !== 'http:' && url.protocol !== 'https:') {
-        throw new UrlValidationError(`Invalid protocol: ${url.protocol}. Only http: and https: are allowed`, urlString);
+        throw new UrlValidationError(`Invalid protocol: ${url.protocol}. Only http: and https: are allowed`, trimmedUrl);
+    }
+    // Block URLs with credentials (user:pass@host)
+    if (url.username || url.password) {
+        throw new UrlValidationError('URLs with embedded credentials are not allowed', trimmedUrl);
     }
     const hostname = url.hostname.toLowerCase();
+    // Block empty hostname
+    if (!hostname) {
+        throw new UrlValidationError('URL must have a valid hostname', trimmedUrl);
+    }
     // Block known internal/metadata hosts
     if (BLOCKED_HOSTS.has(hostname)) {
-        throw new UrlValidationError(`Blocked host: ${hostname}. Internal hosts are not allowed`, urlString);
+        throw new UrlValidationError(`Blocked host: ${hostname}. Internal hosts are not allowed`, trimmedUrl);
     }
     // Block private IP ranges
     if (isBlockedIp(hostname)) {
-        throw new UrlValidationError(`Blocked IP range: ${hostname}. Private IPs are not allowed`, urlString);
+        throw new UrlValidationError(`Blocked IP range: ${hostname}. Private IPs are not allowed`, trimmedUrl);
+    }
+    // Block hostnames that look like they might resolve to internal addresses
+    if (hostname.endsWith('.local') || hostname.endsWith('.internal')) {
+        throw new UrlValidationError(`Blocked hostname pattern: ${hostname}. Internal domain suffixes are not allowed`, trimmedUrl);
     }
     return url.href;
 }

package/dist/utils/url-validator.js.map

@@ -1 +1 @@
-{"version":3,"file":"url-validator.js","sourceRoot":"","sources":["../../src/utils/url-validator.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,kBAAkB,EAAE,MAAM,wBAAwB,CAAC;AAE5D,wCAAwC;AACxC,MAAM,aAAa,GAAG,IAAI,GAAG,CAAC;IAC5B,WAAW;IACX,WAAW;IACX,SAAS;IACT,KAAK;IACL,iBAAiB,EAAE,wBAAwB;IAC3C,0BAA0B,EAAE,eAAe;IAC3C,oBAAoB,EAAE,iBAAiB;CACxC,CAAC,CAAC;AAEH,yCAAyC;AACzC,MAAM,mBAAmB,GAAsB;IAC7C,OAAO,EAAE,kBAAkB;IAC3B,4BAA4B,EAAE,kBAAkB;IAChD,aAAa,EAAE,kBAAkB;IACjC,QAAQ,EAAE,WAAW;IACrB,MAAM,EAAE,kBAAkB;IAC1B,aAAa,EAAE,aAAa;IAC5B,SAAS,EAAE,oBAAoB;IAC/B,SAAS,EAAE,kBAAkB;CAC9B,CAAC;AAEF;;GAEG;AACH,SAAS,WAAW,CAAC,QAAgB;IACnC,OAAO,mBAAmB,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC;AACvE,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,uBAAuB,CAAC,SAAiB;IACvD,IAAI,GAAQ,CAAC;IAEb,IAAI,CAAC;QACH,GAAG,GAAG,IAAI,GAAG,CAAC,SAAS,CAAC,CAAC;IAC3B,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,kBAAkB,CAAC,oBAAoB,EAAE,SAAS,CAAC,CAAC;IAChE,CAAC;IAED,+BAA+B;IAC/B,IAAI,GAAG,CAAC,QAAQ,KAAK,OAAO,IAAI,GAAG,CAAC,QAAQ,KAAK,QAAQ,EAAE,CAAC;QAC1D,MAAM,IAAI,kBAAkB,CAC1B,qBAAqB,GAAG,CAAC,QAAQ,qCAAqC,EACtE,SAAS,CACV,CAAC;IACJ,CAAC;IAED,MAAM,QAAQ,GAAG,GAAG,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC;IAE5C,sCAAsC;IACtC,IAAI,aAAa,CAAC,GAAG,CAAC,QAAQ,CAAC,EAAE,CAAC;QAChC,MAAM,IAAI,kBAAkB,CAC1B,iBAAiB,QAAQ,kCAAkC,EAC3D,SAAS,CACV,CAAC;IACJ,CAAC;IAED,0BAA0B;IAC1B,IAAI,WAAW,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC1B,MAAM,IAAI,kBAAkB,CAC1B,qBAAqB,QAAQ,+BAA+B,EAC5D,SAAS,CACV,CAAC;IACJ,CAAC;IAED,OAAO,GAAG,CAAC,IAAI,CAAC;AAClB,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,aAAa,CAAC,GAAW,EAAE,OAAe;IACxD,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC;QACrC,MAAM,UAAU,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC;QACpC,OAAO,MAAM,CAAC,QAAQ,KAAK,UAAU,CAAC,QAAQ,CAAC;IACjD,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC"}
+{"version":3,"file":"url-validator.js","sourceRoot":"","sources":["../../src/utils/url-validator.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,kBAAkB,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AAE7E,4CAA4C;AAC5C,MAAM,cAAc,GAAG,IAAI,CAAC;AAE5B,wCAAwC;AACxC,MAAM,aAAa,GAAG,IAAI,GAAG,CAAC;IAC5B,WAAW;IACX,WAAW;IACX,SAAS;IACT,KAAK;IACL,iBAAiB,EAAE,wBAAwB;IAC3C,0BAA0B,EAAE,eAAe;IAC3C,oBAAoB,EAAE,iBAAiB;IACvC,iBAAiB,EAAE,yBAAyB;IAC5C,eAAe,EAAE,iCAAiC;CACnD,CAAC,CAAC;AAEH,yCAAyC;AACzC,MAAM,mBAAmB,GAAsB;IAC7C,OAAO,EAAE,kBAAkB;IAC3B,4BAA4B,EAAE,kBAAkB;IAChD,aAAa,EAAE,kBAAkB;IACjC,QAAQ,EAAE,WAAW;IACrB,MAAM,EAAE,kBAAkB;IAC1B,aAAa,EAAE,aAAa;IAC5B,SAAS,EAAE,oBAAoB;IAC/B,SAAS,EAAE,kBAAkB;IAC7B,eAAe,EAAE,4BAA4B;IAC7C,cAAc,EAAE,2BAA2B;IAC3C,mCAAmC,EAAE,2BAA2B;IAChE,oBAAoB,EAAE,2BAA2B;CAClD,CAAC;AAEF;;GAEG;AACH,SAAS,WAAW,CAAC,QAAgB;IACnC,OAAO,mBAAmB,CAAC,IAAI,CAAC,CAAC,OAAO,EAAE,EAAE,CAAC,OAAO,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC;AACvE,CAAC;AAED;;;;GAIG;AACH,MAAM,UAAU,uBAAuB,CAAC,SAAiB;IACvD,2CAA2C;IAC3C,IAAI,CAAC,SAAS,IAAI,OAAO,SAAS,KAAK,QAAQ,EAAE,CAAC;QAChD,MAAM,IAAI,eAAe,CAAC,iBAAiB,CAAC,CAAC;IAC/C,CAAC;IAED,MAAM,UAAU,GAAG,SAAS,CAAC,IAAI,EAAE,CAAC;IACpC,IAAI,CAAC,UAAU,EAAE,CAAC;QAChB,MAAM,IAAI,eAAe,CAAC,qBAAqB,CAAC,CAAC;IACnD,CAAC;IAED,kCAAkC;IAClC,IAAI,UAAU,CAAC,MAAM,GAAG,cAAc,EAAE,CAAC;QACvC,MAAM,IAAI,eAAe,CACvB,iCAAiC,cAAc,aAAa,EAC5D,EAAE,MAAM,EAAE,UAAU,CAAC,MAAM,EAAE,SAAS,EAAE,cAAc,EAAE,CACzD,CAAC;IACJ,CAAC;IAED,IAAI,GAAQ,CAAC;IAEb,IAAI,CAAC;QACH,GAAG,GAAG,IAAI,GAAG,CAAC,UAAU,CAAC,CAAC;IAC5B,CAAC;IAAC,MAAM,CAAC;QACP,MAAM,IAAI,kBAAkB,CAAC,oBAAoB,EAAE,UAAU,CAAC,CAAC;IACjE,CAAC;IAED,+BAA+B;IAC/B,IAAI,GAAG,CAAC,QAAQ,KAAK,OAAO,IAAI,GAAG,CAAC,QAAQ,KAAK,QAAQ,EAAE,CAAC;QAC1D,MAAM,IAAI,kBAAkB,CAC1B,qBAAqB,GAAG,CAAC,QAAQ,qCAAqC,EACtE,UAAU,CACX,CAAC;IACJ,CAAC;IAED,+CAA+C;IAC/C,IAAI,GAAG,CAAC,QAAQ,IAAI,GAAG,CAAC,QAAQ,EAAE,CAAC;QACjC,MAAM,IAAI,kBAAkB,CAC1B,gDAAgD,EAChD,UAAU,CACX,CAAC;IACJ,CAAC;IAED,MAAM,QAAQ,GAAG,GAAG,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC;IAE5C,uBAAuB;IACvB,IAAI,CAAC,QAAQ,EAAE,CAAC;QACd,MAAM,IAAI,kBAAkB,CAAC,gCAAgC,EAAE,UAAU,CAAC,CAAC;IAC7E,CAAC;IAED,sCAAsC;IACtC,IAAI,aAAa,CAAC,GAAG,CAAC,QAAQ,CAAC,EAAE,CAAC;QAChC,MAAM,IAAI,kBAAkB,CAC1B,iBAAiB,QAAQ,kCAAkC,EAC3D,UAAU,CACX,CAAC;IACJ,CAAC;IAED,0BAA0B;IAC1B,IAAI,WAAW,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC1B,MAAM,IAAI,kBAAkB,CAC1B,qBAAqB,QAAQ,+BAA+B,EAC5D,UAAU,CACX,CAAC;IACJ,CAAC;IAED,0EAA0E;IAC1E,IAAI,QAAQ,CAAC,QAAQ,CAAC,QAAQ,CAAC,IAAI,QAAQ,CAAC,QAAQ,CAAC,WAAW,CAAC,EAAE,CAAC;QAClE,MAAM,IAAI,kBAAkB,CAC1B,6BAA6B,QAAQ,4CAA4C,EACjF,UAAU,CACX,CAAC;IACJ,CAAC;IAED,OAAO,GAAG,CAAC,IAAI,CAAC;AAClB,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,aAAa,CAAC,GAAW,EAAE,OAAe;IACxD,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,IAAI,GAAG,CAAC,GAAG,EAAE,OAAO,CAAC,CAAC;QACrC,MAAM,UAAU,GAAG,IAAI,GAAG,CAAC,OAAO,CAAC,CAAC;QACpC,OAAO,MAAM,CAAC,QAAQ,KAAK,UAAU,CAAC,QAAQ,CAAC;IACjD,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC"}