@agenticmail/enterprise 0.5.327 → 0.5.328

package/src/agent-tools/tools/enterprise-documents.ts

@@ -1,896 +0,0 @@
-/**
- * AgenticMail Agent Tools — Enterprise Documents
- *
- * Document generation, parsing, and conversion tools for AI agents.
- * Generates valid PDFs and DOCX files using raw format construction,
- * extracts invoice data with regex patterns, and converts between formats.
- */
-
-import fs from 'node:fs/promises';
-import path from 'node:path';
-import { execFile } from 'node:child_process';
-import zlib from 'node:zlib';
-import { promisify } from 'node:util';
-import type { AnyAgentTool, ToolCreationOptions } from '../types.js';
-import { readStringParam, jsonResult, textResult, errorResult } from '../common.js';
-
-var deflateRaw = promisify(zlib.deflateRaw);
-
-// --- Minimal PDF Writer ---
-
-function buildPdf(textLines: string[], title?: string): Buffer {
-  var offsets: number[] = [];
-  var body = '';
-  var objNum = 0;
-
-  function addObj(content: string): number {
-    objNum++;
-    offsets.push(Buffer.byteLength(body, 'binary'));
-    body += objNum + ' 0 obj\n' + content + '\nendobj\n';
-    return objNum;
-  }
-
-  // Catalog (obj 1)
-  addObj('<< /Type /Catalog /Pages 2 0 R >>');
-
-  // Pages (obj 2)
-  addObj('<< /Type /Pages /Kids [3 0 R] /Count 1 >>');
-
-  // Build text stream
-  var fontSize = 12;
-  var leading = 14;
-  var margin = 72;
-  var pageWidth = 612;
-  var pageHeight = 792;
-  var usableWidth = pageWidth - 2 * margin;
-
-  var stream = 'BT\n/F1 ' + fontSize + ' Tf\n' + margin + ' ' + (pageHeight - margin) + ' Td\n' + leading + ' TL\n';
-  if (title) {
-    stream += '/F1 16 Tf\n(' + escapePdfString(title) + ') Tj\nT*\n/F1 ' + fontSize + ' Tf\nT*\n';
-  }
-  for (var line of textLines) {
-    // Wrap long lines
-    var wrapped = wrapText(line, usableWidth, fontSize * 0.6);
-    for (var wl of wrapped) {
-      stream += '(' + escapePdfString(wl) + ') Tj\nT*\n';
-    }
-  }
-  stream += 'ET';
-
-  // Content stream (obj 4)
-  addObj('<< /Length ' + stream.length + ' >>\nstream\n' + stream + '\nendstream');
-
-  // Page (obj 3) — inserted before stream in numbering, but we adjust
-  // Re-do: we need page=3 and stream=4
-  // Actually we already added catalog=1, pages=2, and streamObj got 3
-  // Let's restructure: add page after stream
-
-  // Reset and rebuild properly
-  body = '';
-  offsets = [];
-  objNum = 0;
-
-  // 1: Catalog
-  addObj('<< /Type /Catalog /Pages 2 0 R >>');
-  // 2: Pages
-  addObj('<< /Type /Pages /Kids [3 0 R] /Count 1 >>');
-  // 3: Page
-  addObj('<< /Type /Page /Parent 2 0 R /MediaBox [0 0 ' + pageWidth + ' ' + pageHeight + '] /Contents 5 0 R /Resources << /Font << /F1 4 0 R >> >> >>');
-  // 4: Font
-  addObj('<< /Type /Font /Subtype /Type1 /BaseFont /Helvetica >>');
-  // 5: Content stream
-  addObj('<< /Length ' + stream.length + ' >>\nstream\n' + stream + '\nendstream');
-
-  // Build PDF
-  var header = '%PDF-1.4\n%\xE2\xE3\xCF\xD3\n';
-  var xrefOffset = Buffer.byteLength(header, 'binary') + Buffer.byteLength(body, 'binary');
-
-  var xref = 'xref\n0 ' + (objNum + 1) + '\n0000000000 65535 f \n';
-  for (var i = 0; i < offsets.length; i++) {
-    var off = Buffer.byteLength(header, 'binary') + offsets[i];
-    xref += String(off).padStart(10, '0') + ' 00000 n \n';
-  }
-
-  var trailer = 'trailer\n<< /Size ' + (objNum + 1) + ' /Root 1 0 R >>\nstartxref\n' + xrefOffset + '\n%%EOF\n';
-
-  return Buffer.from(header + body + xref + trailer, 'binary');
-}
-
-function escapePdfString(str: string): string {
-  return str
-    .replace(/\\/g, '\\\\')
-    .replace(/\(/g, '\\(')
-    .replace(/\)/g, '\\)')
-    .replace(/\r/g, '\\r');
-}
-
-function wrapText(text: string, maxWidth: number, charWidth: number): string[] {
-  var maxChars = Math.floor(maxWidth / charWidth);
-  if (text.length <= maxChars) return [text];
-  var lines: string[] = [];
-  var remaining = text;
-  while (remaining.length > 0) {
-    if (remaining.length <= maxChars) {
-      lines.push(remaining);
-      break;
-    }
-    var breakAt = remaining.lastIndexOf(' ', maxChars);
-    if (breakAt <= 0) breakAt = maxChars;
-    lines.push(remaining.slice(0, breakAt));
-    remaining = remaining.slice(breakAt).replace(/^\s+/, '');
-  }
-  return lines;
-}
-
-// --- Minimal DOCX Builder ---
-
-async function buildDocx(paragraphs: string[], title?: string): Promise<Buffer> {
-  // DOCX is a ZIP of XML files. We build a minimal one using raw ZIP construction.
-  var contentXml = '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
-    + '<w:document xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main"'
-    + ' xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships">'
-    + '<w:body>';
-
-  if (title) {
-    contentXml += '<w:p><w:pPr><w:pStyle w:val="Heading1"/></w:pPr>'
-      + '<w:r><w:rPr><w:b/><w:sz w:val="32"/></w:rPr>'
-      + '<w:t>' + escapeXml(title) + '</w:t></w:r></w:p>';
-  }
-
-  for (var para of paragraphs) {
-    contentXml += '<w:p><w:r><w:t xml:space="preserve">' + escapeXml(para) + '</w:t></w:r></w:p>';
-  }
-
-  contentXml += '</w:body></w:document>';
-
-  var contentTypesXml = '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
-    + '<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types">'
-    + '<Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/>'
-    + '<Default Extension="xml" ContentType="application/xml"/>'
-    + '<Override PartName="/word/document.xml" ContentType="application/vnd.openxmlformats-officedocument.wordprocessingml.document.main+xml"/>'
-    + '</Types>';
-
-  var relsXml = '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
-    + '<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">'
-    + '<Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="word/document.xml"/>'
-    + '</Relationships>';
-
-  var wordRelsXml = '<?xml version="1.0" encoding="UTF-8" standalone="yes"?>'
-    + '<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">'
-    + '</Relationships>';
-
-  var files: Array<{ name: string; content: Buffer }> = [
-    { name: '[Content_Types].xml', content: Buffer.from(contentTypesXml, 'utf-8') },
-    { name: '_rels/.rels', content: Buffer.from(relsXml, 'utf-8') },
-    { name: 'word/document.xml', content: Buffer.from(contentXml, 'utf-8') },
-    { name: 'word/_rels/document.xml.rels', content: Buffer.from(wordRelsXml, 'utf-8') },
-  ];
-
-  return buildZip(files);
-}
-
-function escapeXml(str: string): string {
-  return str
-    .replace(/&/g, '&amp;')
-    .replace(/</g, '&lt;')
-    .replace(/>/g, '&gt;')
-    .replace(/"/g, '&quot;')
-    .replace(/'/g, '&apos;');
-}
-
-async function buildZip(files: Array<{ name: string; content: Buffer }>): Promise<Buffer> {
-  var localHeaders: Buffer[] = [];
-  var centralHeaders: Buffer[] = [];
-  var offset = 0;
-
-  for (var file of files) {
-    var nameBuffer = Buffer.from(file.name, 'utf-8');
-    var compressed = await deflateRaw(file.content);
-    var crc = crc32(file.content);
-
-    // Local file header
-    var local = Buffer.alloc(30 + nameBuffer.length);
-    local.writeUInt32LE(0x04034b50, 0);   // signature
-    local.writeUInt16LE(20, 4);            // version needed
-    local.writeUInt16LE(0, 6);             // flags
-    local.writeUInt16LE(8, 8);             // compression: deflate
-    local.writeUInt16LE(0, 10);            // mod time
-    local.writeUInt16LE(0, 12);            // mod date
-    local.writeUInt32LE(crc, 14);          // CRC-32
-    local.writeUInt32LE(compressed.length, 18);  // compressed size
-    local.writeUInt32LE(file.content.length, 22); // uncompressed size
-    local.writeUInt16LE(nameBuffer.length, 26);   // name length
-    local.writeUInt16LE(0, 28);            // extra field length
-    nameBuffer.copy(local, 30);
-
-    localHeaders.push(Buffer.concat([local, compressed]));
-
-    // Central directory header
-    var central = Buffer.alloc(46 + nameBuffer.length);
-    central.writeUInt32LE(0x02014b50, 0);  // signature
-    central.writeUInt16LE(20, 4);          // version made by
-    central.writeUInt16LE(20, 6);          // version needed
-    central.writeUInt16LE(0, 8);           // flags
-    central.writeUInt16LE(8, 10);          // compression
-    central.writeUInt16LE(0, 12);          // mod time
-    central.writeUInt16LE(0, 14);          // mod date
-    central.writeUInt32LE(crc, 16);
-    central.writeUInt32LE(compressed.length, 20);
-    central.writeUInt32LE(file.content.length, 24);
-    central.writeUInt16LE(nameBuffer.length, 28);
-    central.writeUInt16LE(0, 30);          // extra field length
-    central.writeUInt16LE(0, 32);          // comment length
-    central.writeUInt16LE(0, 34);          // disk number
-    central.writeUInt16LE(0, 36);          // internal attributes
-    central.writeUInt32LE(0, 38);          // external attributes
-    central.writeUInt32LE(offset, 42);     // local header offset
-    nameBuffer.copy(central, 46);
-
-    centralHeaders.push(central);
-    offset += local.length + compressed.length;
-  }
-
-  var centralOffset = offset;
-  var centralSize = centralHeaders.reduce(function(sum, b) { return sum + b.length; }, 0);
-
-  // End of central directory
-  var eocd = Buffer.alloc(22);
-  eocd.writeUInt32LE(0x06054b50, 0);
-  eocd.writeUInt16LE(0, 4);              // disk number
-  eocd.writeUInt16LE(0, 6);              // central dir disk
-  eocd.writeUInt16LE(files.length, 8);   // entries on disk
-  eocd.writeUInt16LE(files.length, 10);  // total entries
-  eocd.writeUInt32LE(centralSize, 12);
-  eocd.writeUInt32LE(centralOffset, 16);
-  eocd.writeUInt16LE(0, 20);             // comment length
-
-  return Buffer.concat([...localHeaders, ...centralHeaders, eocd]);
-}
-
-function crc32(buf: Buffer): number {
-  var table: number[] = [];
-  for (var n = 0; n < 256; n++) {
-    var c = n;
-    for (var k = 0; k < 8; k++) {
-      if (c & 1) c = 0xEDB88320 ^ (c >>> 1);
-      else c = c >>> 1;
-    }
-    table[n] = c;
-  }
-  var crc = 0xFFFFFFFF;
-  for (var i = 0; i < buf.length; i++) {
-    crc = table[(crc ^ buf[i]) & 0xFF] ^ (crc >>> 8);
-  }
-  return (crc ^ 0xFFFFFFFF) >>> 0;
-}
-
-export function createDocumentTools(options?: ToolCreationOptions): AnyAgentTool[] {
-
-  var entDocGeneratePdf: AnyAgentTool = {
-    name: 'ent_doc_generate_pdf',
-    label: 'Generate PDF',
-    description: 'Generate a PDF document from plain text content.',
-    category: 'file',
-    risk: 'medium',
-    parameters: {
-      type: 'object',
-      properties: {
-        output_path: { type: 'string', description: 'Path to write the PDF file.' },
-        content: { type: 'string', description: 'Text content for the PDF body (one paragraph per line).' },
-        title: { type: 'string', description: 'Optional document title.' },
-      },
-      required: ['output_path', 'content'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var outputPath = readStringParam(params, 'output_path', { required: true });
-      var content = readStringParam(params, 'content', { required: true, trim: false });
-      var title = readStringParam(params, 'title');
-
-      if (!path.isAbsolute(outputPath) && options?.workspaceDir) {
-        outputPath = path.resolve(options.workspaceDir, outputPath);
-      }
-
-      try {
-        var lines = content.split('\n');
-        var pdfBuffer = buildPdf(lines, title);
-        var dir = path.dirname(outputPath);
-        await fs.mkdir(dir, { recursive: true });
-        await fs.writeFile(outputPath, pdfBuffer);
-        return textResult('Generated PDF: ' + outputPath + ' (' + pdfBuffer.length + ' bytes, ' + lines.length + ' lines)');
-      } catch (err: any) {
-        return errorResult('PDF generation failed: ' + (err.message || String(err)));
-      }
-    },
-  };
-
-  var entDocGenerateDocx: AnyAgentTool = {
-    name: 'ent_doc_generate_docx',
-    label: 'Generate DOCX',
-    description: 'Generate a Microsoft Word DOCX document from plain text content.',
-    category: 'file',
-    risk: 'medium',
-    parameters: {
-      type: 'object',
-      properties: {
-        output_path: { type: 'string', description: 'Path to write the DOCX file.' },
-        content: { type: 'string', description: 'Text content (one paragraph per line).' },
-        title: { type: 'string', description: 'Optional document title.' },
-      },
-      required: ['output_path', 'content'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var outputPath = readStringParam(params, 'output_path', { required: true });
-      var content = readStringParam(params, 'content', { required: true, trim: false });
-      var title = readStringParam(params, 'title');
-
-      if (!path.isAbsolute(outputPath) && options?.workspaceDir) {
-        outputPath = path.resolve(options.workspaceDir, outputPath);
-      }
-
-      try {
-        var paragraphs = content.split('\n');
-        var docxBuffer = await buildDocx(paragraphs, title);
-        var dir = path.dirname(outputPath);
-        await fs.mkdir(dir, { recursive: true });
-        await fs.writeFile(outputPath, docxBuffer);
-        return textResult('Generated DOCX: ' + outputPath + ' (' + docxBuffer.length + ' bytes, ' + paragraphs.length + ' paragraphs)');
-      } catch (err: any) {
-        return errorResult('DOCX generation failed: ' + (err.message || String(err)));
-      }
-    },
-  };
-
-  var entDocOcr: AnyAgentTool = {
-    name: 'ent_doc_ocr',
-    label: 'OCR Image',
-    description: 'Extract text from an image file using OCR.',
-    category: 'utility',
-    risk: 'low',
-    parameters: {
-      type: 'object',
-      properties: {
-        file_path: { type: 'string', description: 'Path to the image file (PNG, JPG, TIFF, BMP).' },
-        language: { type: 'string', description: 'OCR language code (default "eng").' },
-      },
-      required: ['file_path'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var filePath = readStringParam(params, 'file_path', { required: true });
-      var language = readStringParam(params, 'language') || 'eng';
-
-      if (!path.isAbsolute(filePath) && options?.workspaceDir) {
-        filePath = path.resolve(options.workspaceDir, filePath);
-      }
-
-      try {
-        await fs.access(filePath);
-      } catch {
-        return errorResult('File not found: ' + filePath);
-      }
-
-      // Try tesseract CLI
-      try {
-        var result = await new Promise<string>(function(resolve, reject) {
-          execFile('tesseract', [filePath, 'stdout', '-l', language], { timeout: 30000 }, function(err, stdout, _stderr) {
-            if (err) reject(err);
-            else resolve(stdout);
-          });
-        });
-        var text = result.trim();
-        if (!text) return textResult('OCR completed but no text was detected in the image.');
-        return jsonResult({ file: filePath, extractedText: text, charCount: text.length, lineCount: text.split('\n').length });
-      } catch {
-        // Tesseract not available — return metadata and suggestion
-        var stat = await fs.stat(filePath);
-        var ext = path.extname(filePath).toLowerCase();
-        return jsonResult({
-          file: filePath,
-          size: stat.size,
-          format: ext,
-          message: 'Tesseract OCR is not installed. Install it with: brew install tesseract (macOS) or apt-get install tesseract-ocr (Linux). Alternatively, use the vision/read tool to view the image directly.',
-        });
-      }
-    },
-  };
-
-  var entDocParseInvoice: AnyAgentTool = {
-    name: 'ent_doc_parse_invoice',
-    label: 'Parse Invoice',
-    description: 'Parse structured text to extract invoice fields: invoice number, date, vendor...',
-    category: 'utility',
-    risk: 'low',
-    parameters: {
-      type: 'object',
-      properties: {
-        file_path: { type: 'string', description: 'Path to a text file containing invoice content.' },
-        content: { type: 'string', description: 'Alternatively, provide invoice text content directly.' },
-      },
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var filePath = readStringParam(params, 'file_path');
-      var content = readStringParam(params, 'content', { trim: false });
-
-      if (!content && filePath) {
-        var resolved = filePath;
-        if (!path.isAbsolute(resolved) && options?.workspaceDir) {
-          resolved = path.resolve(options.workspaceDir, resolved);
-        }
-        try {
-          content = await fs.readFile(resolved, 'utf-8');
-        } catch (err: any) {
-          return errorResult('Failed to read file: ' + (err.message || String(err)));
-        }
-      }
-
-      if (!content) {
-        return errorResult('Provide either file_path or content parameter.');
-      }
-
-      // Extract invoice fields using regex patterns
-      var invoice: Record<string, any> = {};
-
-      // Invoice number
-      var invNumMatch = content.match(/(?:invoice|inv)[\s#:]*([A-Z0-9\-]+)/i);
-      if (invNumMatch) invoice.invoiceNumber = invNumMatch[1];
-
-      // Date patterns
-      var dateMatch = content.match(/(?:date|issued|invoice date)[:\s]*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{2,4}|\w+ \d{1,2},?\s*\d{4}|\d{4}-\d{2}-\d{2})/i);
-      if (dateMatch) invoice.date = dateMatch[1].trim();
-
-      // Due date
-      var dueDateMatch = content.match(/(?:due date|payment due|due)[:\s]*(\d{1,2}[\/\-]\d{1,2}[\/\-]\d{2,4}|\w+ \d{1,2},?\s*\d{4}|\d{4}-\d{2}-\d{2})/i);
-      if (dueDateMatch) invoice.dueDate = dueDateMatch[1].trim();
-
-      // Vendor / From
-      var vendorMatch = content.match(/(?:from|vendor|seller|billed by|company)[:\s]*([^\n]+)/i);
-      if (vendorMatch) invoice.vendor = vendorMatch[1].trim();
-
-      // Bill to / Customer
-      var customerMatch = content.match(/(?:to|bill to|customer|client|billed to)[:\s]*([^\n]+)/i);
-      if (customerMatch) invoice.customer = customerMatch[1].trim();
-
-      // Total
-      var totalMatch = content.match(/(?:total|amount due|balance due|grand total)[:\s]*([\$\u00A3\u20AC]?\s?[\d,]+\.?\d{0,2})/i);
-      if (totalMatch) invoice.total = totalMatch[1].trim();
-
-      // Subtotal
-      var subtotalMatch = content.match(/(?:subtotal|sub-total|sub total)[:\s]*([\$\u00A3\u20AC]?\s?[\d,]+\.?\d{0,2})/i);
-      if (subtotalMatch) invoice.subtotal = subtotalMatch[1].trim();
-
-      // Tax
-      var taxMatch = content.match(/(?:tax|vat|gst|hst)[:\s]*([\$\u00A3\u20AC]?\s?[\d,]+\.?\d{0,2})/i);
-      if (taxMatch) invoice.tax = taxMatch[1].trim();
-
-      // Line items: look for lines with description + amount pattern
-      var lineItems: Array<{ description: string; amount: string }> = [];
-      var lines = content.split('\n');
-      for (var line of lines) {
-        var lineMatch = line.match(/^(.+?)\s+([\$\u00A3\u20AC]?\s?[\d,]+\.\d{2})\s*$/);
-        if (lineMatch) {
-          var desc = lineMatch[1].trim();
-          // Skip header/total lines
-          if (!/^(subtotal|total|tax|vat|amount|balance|due|date|invoice)/i.test(desc)) {
-            lineItems.push({ description: desc, amount: lineMatch[2].trim() });
-          }
-        }
-      }
-      if (lineItems.length > 0) invoice.lineItems = lineItems;
-
-      var fieldsFound = Object.keys(invoice).length;
-      if (fieldsFound === 0) {
-        return textResult('No invoice fields could be extracted. The content may not be in a recognized invoice format.');
-      }
-
-      invoice._fieldsExtracted = fieldsFound;
-      return jsonResult(invoice);
-    },
-  };
-
-  var entDocConvert: AnyAgentTool = {
-    name: 'ent_doc_convert',
-    label: 'Convert Document',
-    description: 'Convert between document formats: markdown to HTML, HTML to plain text, JSON ...',
-    category: 'utility',
-    risk: 'medium',
-    parameters: {
-      type: 'object',
-      properties: {
-        input_path: { type: 'string', description: 'Path to the input file.' },
-        output_path: { type: 'string', description: 'Path to write the output file.' },
-        conversion: { type: 'string', description: 'Conversion type.', enum: ['md_to_html', 'html_to_text', 'json_to_csv', 'text_to_md'] },
-      },
-      required: ['input_path', 'output_path', 'conversion'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var inputPath = readStringParam(params, 'input_path', { required: true });
-      var outputPath = readStringParam(params, 'output_path', { required: true });
-      var conversion = readStringParam(params, 'conversion', { required: true });
-
-      if (!path.isAbsolute(inputPath) && options?.workspaceDir) {
-        inputPath = path.resolve(options.workspaceDir, inputPath);
-      }
-      if (!path.isAbsolute(outputPath) && options?.workspaceDir) {
-        outputPath = path.resolve(options.workspaceDir, outputPath);
-      }
-
-      try {
-        var content = await fs.readFile(inputPath, 'utf-8');
-        var output: string;
-
-        switch (conversion) {
-          case 'md_to_html': {
-            // Simple markdown to HTML conversion
-            output = content
-              .replace(/^### (.+)$/gm, '<h3>$1</h3>')
-              .replace(/^## (.+)$/gm, '<h2>$1</h2>')
-              .replace(/^# (.+)$/gm, '<h1>$1</h1>')
-              .replace(/\*\*(.+?)\*\*/g, '<strong>$1</strong>')
-              .replace(/\*(.+?)\*/g, '<em>$1</em>')
-              .replace(/`(.+?)`/g, '<code>$1</code>')
-              .replace(/^\- (.+)$/gm, '<li>$1</li>')
-              .replace(/^\d+\. (.+)$/gm, '<li>$1</li>')
-              .replace(/\[(.+?)\]\((.+?)\)/g, '<a href="$2">$1</a>')
-              .replace(/^(?!<[hlo])(.*[^\n])$/gm, '<p>$1</p>')
-              .replace(/\n{2,}/g, '\n');
-            output = '<!DOCTYPE html>\n<html>\n<head><meta charset="utf-8"></head>\n<body>\n' + output + '\n</body>\n</html>';
-            break;
-          }
-          case 'html_to_text': {
-            output = content
-              .replace(/<script[^>]*>[\s\S]*?<\/script>/gi, '')
-              .replace(/<style[^>]*>[\s\S]*?<\/style>/gi, '')
-              .replace(/<br\s*\/?>/gi, '\n')
-              .replace(/<\/p>/gi, '\n\n')
-              .replace(/<\/div>/gi, '\n')
-              .replace(/<\/h[1-6]>/gi, '\n\n')
-              .replace(/<\/li>/gi, '\n')
-              .replace(/<li[^>]*>/gi, '  - ')
-              .replace(/<[^>]+>/g, '')
-              .replace(/&amp;/g, '&')
-              .replace(/&lt;/g, '<')
-              .replace(/&gt;/g, '>')
-              .replace(/&quot;/g, '"')
-              .replace(/&#39;/g, "'")
-              .replace(/&nbsp;/g, ' ')
-              .replace(/\n{3,}/g, '\n\n')
-              .trim();
-            break;
-          }
-          case 'json_to_csv': {
-            var data = JSON.parse(content);
-            if (!Array.isArray(data)) {
-              return errorResult('JSON input must be an array of objects for CSV conversion.');
-            }
-            if (data.length === 0) {
-              output = '';
-              break;
-            }
-            var headers = Object.keys(data[0]);
-            var csvLines = [headers.map(function(h) {
-              return h.indexOf(',') >= 0 || h.indexOf('"') >= 0 ? '"' + h.replace(/"/g, '""') + '"' : h;
-            }).join(',')];
-            for (var row of data) {
-              csvLines.push(headers.map(function(h) {
-                var val = row[h] != null ? String(row[h]) : '';
-                return val.indexOf(',') >= 0 || val.indexOf('"') >= 0 || val.indexOf('\n') >= 0
-                  ? '"' + val.replace(/"/g, '""') + '"' : val;
-              }).join(','));
-            }
-            output = csvLines.join('\n') + '\n';
-            break;
-          }
-          case 'text_to_md': {
-            // Convert plain text to markdown: detect paragraphs, lists, headings
-            var lines = content.split('\n');
-            var mdLines: string[] = [];
-            for (var line of lines) {
-              var trimmed = line.trim();
-              if (!trimmed) {
-                mdLines.push('');
-                continue;
-              }
-              // Lines that are ALL CAPS could be headings
-              if (trimmed.length > 3 && trimmed === trimmed.toUpperCase() && /^[A-Z\s]+$/.test(trimmed)) {
-                mdLines.push('## ' + trimmed.charAt(0) + trimmed.slice(1).toLowerCase());
-              }
-              // Lines starting with - or * are already list-like
-              else if (/^[\-\*]\s/.test(trimmed)) {
-                mdLines.push(trimmed);
-              }
-              // Lines starting with a number + period are ordered lists
-              else if (/^\d+[\.\)]\s/.test(trimmed)) {
-                mdLines.push(trimmed);
-              }
-              else {
-                mdLines.push(trimmed);
-              }
-            }
-            output = mdLines.join('\n');
-            break;
-          }
-          default:
-            return errorResult('Unknown conversion type: ' + conversion);
-        }
-
-        var dir = path.dirname(outputPath);
-        await fs.mkdir(dir, { recursive: true });
-        await fs.writeFile(outputPath, output, 'utf-8');
-        return textResult('Converted ' + conversion + ': ' + inputPath + ' -> ' + outputPath + ' (' + output.length + ' chars)');
-      } catch (err: any) {
-        return errorResult('Conversion failed: ' + (err.message || String(err)));
-      }
-    },
-  };
-
-  var entDocExtractTables: AnyAgentTool = {
-    name: 'ent_doc_extract_tables',
-    label: 'Extract Tables',
-    description: 'Extract tabular data from HTML files (HTML tables) or text files (aligned col...',
-    category: 'utility',
-    risk: 'low',
-    parameters: {
-      type: 'object',
-      properties: {
-        file_path: { type: 'string', description: 'Path to the HTML or text file.' },
-      },
-      required: ['file_path'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var filePath = readStringParam(params, 'file_path', { required: true });
-
-      if (!path.isAbsolute(filePath) && options?.workspaceDir) {
-        filePath = path.resolve(options.workspaceDir, filePath);
-      }
-
-      try {
-        var content = await fs.readFile(filePath, 'utf-8');
-        var ext = path.extname(filePath).toLowerCase();
-        var tables: Array<{ headers: string[]; rows: string[][] }> = [];
-
-        if (ext === '.html' || ext === '.htm' || content.indexOf('<table') >= 0) {
-          // Extract HTML tables
-          var tableRegex = /<table[^>]*>([\s\S]*?)<\/table>/gi;
-          var tableMatch: RegExpExecArray | null;
-          while ((tableMatch = tableRegex.exec(content)) !== null) {
-            var tableHtml = tableMatch[1];
-            var headers: string[] = [];
-            var rows: string[][] = [];
-
-            // Extract headers from <th>
-            var thRegex = /<th[^>]*>([\s\S]*?)<\/th>/gi;
-            var thMatch: RegExpExecArray | null;
-            while ((thMatch = thRegex.exec(tableHtml)) !== null) {
-              headers.push(thMatch[1].replace(/<[^>]+>/g, '').trim());
-            }
-
-            // Extract rows from <tr>
-            var trRegex = /<tr[^>]*>([\s\S]*?)<\/tr>/gi;
-            var trMatch: RegExpExecArray | null;
-            var firstRow = true;
-            while ((trMatch = trRegex.exec(tableHtml)) !== null) {
-              var cells: string[] = [];
-              var tdRegex = /<td[^>]*>([\s\S]*?)<\/td>/gi;
-              var tdMatch: RegExpExecArray | null;
-              while ((tdMatch = tdRegex.exec(trMatch[1])) !== null) {
-                cells.push(tdMatch[1].replace(/<[^>]+>/g, '').trim());
-              }
-              if (cells.length > 0) {
-                if (headers.length === 0 && firstRow) {
-                  headers = cells;
-                } else {
-                  rows.push(cells);
-                }
-              }
-              firstRow = false;
-            }
-
-            if (headers.length > 0 || rows.length > 0) {
-              tables.push({ headers: headers, rows: rows });
-            }
-          }
-        } else {
-          // Try to detect text tables (pipe-delimited or tab-delimited or aligned columns)
-          var lines = content.split('\n').filter(function(l) { return l.trim() !== ''; });
-
-          // Check for pipe-delimited tables (markdown-style)
-          var pipeLines = lines.filter(function(l) { return l.indexOf('|') >= 0; });
-          if (pipeLines.length >= 2) {
-            var dataLines = pipeLines.filter(function(l) { return !/^[\s|:-]+$/.test(l); });
-            if (dataLines.length > 0) {
-              var headers = dataLines[0].split('|').map(function(c) { return c.trim(); }).filter(Boolean);
-              var rows: string[][] = [];
-              for (var i = 1; i < dataLines.length; i++) {
-                var cells = dataLines[i].split('|').map(function(c) { return c.trim(); }).filter(Boolean);
-                if (cells.length > 0) rows.push(cells);
-              }
-              tables.push({ headers: headers, rows: rows });
-            }
-          }
-
-          // Check for tab-delimited
-          if (tables.length === 0) {
-            var tabLines = lines.filter(function(l) { return l.indexOf('\t') >= 0; });
-            if (tabLines.length >= 2) {
-              var headers = tabLines[0].split('\t').map(function(c) { return c.trim(); });
-              var rows: string[][] = [];
-              for (var i = 1; i < tabLines.length; i++) {
-                rows.push(tabLines[i].split('\t').map(function(c) { return c.trim(); }));
-              }
-              tables.push({ headers: headers, rows: rows });
-            }
-          }
-        }
-
-        if (tables.length === 0) {
-          return textResult('No tables found in ' + filePath);
-        }
-
-        return jsonResult({ file: filePath, tableCount: tables.length, tables: tables });
-      } catch (err: any) {
-        return errorResult('Table extraction failed: ' + (err.message || String(err)));
-      }
-    },
-  };
-
-  var entDocMergePdfs: AnyAgentTool = {
-    name: 'ent_doc_merge_pdfs',
-    label: 'Merge PDFs',
-    description: 'Concatenate multiple PDF files into a single PDF.',
-    category: 'file',
-    risk: 'medium',
-    parameters: {
-      type: 'object',
-      properties: {
-        input_files: { type: 'string', description: 'JSON array of PDF file paths to merge, in order.' },
-        output_path: { type: 'string', description: 'Path to write the merged PDF.' },
-      },
-      required: ['input_files', 'output_path'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var inputFilesStr = readStringParam(params, 'input_files', { required: true });
-      var outputPath = readStringParam(params, 'output_path', { required: true });
-
-      if (!path.isAbsolute(outputPath) && options?.workspaceDir) {
-        outputPath = path.resolve(options.workspaceDir, outputPath);
-      }
-
-      try {
-        var inputFiles: string[];
-        try {
-          inputFiles = JSON.parse(inputFilesStr);
-        } catch {
-          return errorResult('Invalid JSON for input_files. Expected a JSON array of file paths.');
-        }
-
-        if (!Array.isArray(inputFiles) || inputFiles.length < 2) {
-          return errorResult('At least 2 PDF files are required for merging.');
-        }
-
-        // Collect all text content from PDFs by re-reading them
-        // For simplicity, we extract text streams and regenerate a single PDF
-        var allLines: string[] = [];
-        for (var i = 0; i < inputFiles.length; i++) {
-          var filePath = inputFiles[i];
-          if (!path.isAbsolute(filePath) && options?.workspaceDir) {
-            filePath = path.resolve(options.workspaceDir, filePath);
-          }
-          try {
-            await fs.access(filePath);
-          } catch {
-            return errorResult('File not found: ' + filePath);
-          }
-
-          var pdfContent = await fs.readFile(filePath);
-          // Extract text between BT and ET markers
-          var pdfStr = pdfContent.toString('binary');
-          var btIdx = pdfStr.indexOf('BT\n');
-          var etIdx = pdfStr.indexOf('\nET', btIdx);
-          if (btIdx >= 0 && etIdx >= 0) {
-            var textBlock = pdfStr.slice(btIdx + 3, etIdx);
-            var textMatches = textBlock.match(/\(([^)]*)\)\s*Tj/g);
-            if (textMatches) {
-              if (i > 0) allLines.push('--- Page ' + (i + 1) + ' ---');
-              for (var tm of textMatches) {
-                var text = tm.replace(/^\(/, '').replace(/\)\s*Tj$/, '')
-                  .replace(/\\\(/g, '(').replace(/\\\)/g, ')').replace(/\\\\/g, '\\');
-                allLines.push(text);
-              }
-            }
-          }
-        }
-
-        var mergedPdf = buildPdf(allLines, undefined);
-        var dir = path.dirname(outputPath);
-        await fs.mkdir(dir, { recursive: true });
-        await fs.writeFile(outputPath, mergedPdf);
-        return textResult('Merged ' + inputFiles.length + ' PDFs into ' + outputPath + ' (' + mergedPdf.length + ' bytes)');
-      } catch (err: any) {
-        return errorResult('PDF merge failed: ' + (err.message || String(err)));
-      }
-    },
-  };
-
-  var entDocFillForm: AnyAgentTool = {
-    name: 'ent_doc_fill_form',
-    label: 'Fill Document Template',
-    description: 'Replace {{placeholder}} patterns in a document template with provided values.',
-    category: 'file',
-    risk: 'medium',
-    parameters: {
-      type: 'object',
-      properties: {
-        template_path: { type: 'string', description: 'Path to the template file with {{placeholder}} patterns.' },
-        output_path: { type: 'string', description: 'Path to write the filled document.' },
-        values: { type: 'string', description: 'JSON object of placeholder names to values (e.g., {"name": "John", "date": "2...' },
-      },
-      required: ['template_path', 'output_path', 'values'],
-    },
-    execute: async function(_toolCallId, args) {
-      var params = args as Record<string, unknown>;
-      var templatePath = readStringParam(params, 'template_path', { required: true });
-      var outputPath = readStringParam(params, 'output_path', { required: true });
-      var valuesStr = readStringParam(params, 'values', { required: true });
-
-      if (!path.isAbsolute(templatePath) && options?.workspaceDir) {
-        templatePath = path.resolve(options.workspaceDir, templatePath);
-      }
-      if (!path.isAbsolute(outputPath) && options?.workspaceDir) {
-        outputPath = path.resolve(options.workspaceDir, outputPath);
-      }
-
-      try {
-        var values: Record<string, string>;
-        try {
-          values = JSON.parse(valuesStr);
-        } catch {
-          return errorResult('Invalid JSON for values parameter.');
-        }
-
-        var template = await fs.readFile(templatePath, 'utf-8');
-        var filled = template;
-        var replacedCount = 0;
-        var missingPlaceholders: string[] = [];
-
-        // Find all placeholders in the template
-        var placeholderRegex = /\{\{(\w+)\}\}/g;
-        var match: RegExpExecArray | null;
-        var foundPlaceholders = new Set<string>();
-        while ((match = placeholderRegex.exec(template)) !== null) {
-          foundPlaceholders.add(match[1]);
-        }
-
-        // Replace each placeholder
-        for (var placeholder of foundPlaceholders) {
-          if (values[placeholder] !== undefined) {
-            var pattern = new RegExp('\\{\\{' + placeholder + '\\}\\}', 'g');
-            filled = filled.replace(pattern, values[placeholder]);
-            replacedCount++;
-          } else {
-            missingPlaceholders.push(placeholder);
-          }
-        }
-
-        var dir = path.dirname(outputPath);
-        await fs.mkdir(dir, { recursive: true });
-        await fs.writeFile(outputPath, filled, 'utf-8');
-
-        var summary = 'Filled ' + replacedCount + ' placeholder(s) in template. Wrote to ' + outputPath;
-        if (missingPlaceholders.length > 0) {
-          summary += '\nWarning: ' + missingPlaceholders.length + ' unfilled placeholder(s): ' + missingPlaceholders.join(', ');
-        }
-        return textResult(summary);
-      } catch (err: any) {
-        return errorResult('Template fill failed: ' + (err.message || String(err)));
-      }
-    },
-  };
-
-  return [entDocGeneratePdf, entDocGenerateDocx, entDocOcr, entDocParseInvoice, entDocConvert, entDocExtractTables, entDocMergePdfs, entDocFillForm];
-}