Publish expanded 23-language matrix across all 20 inventions (460 codebases total)

.gitattributes

@@ -36,3 +36,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 architecture.png filter=lfs diff=lfs merge=lfs -text
 Logo.jpg filter=lfs diff=lfs merge=lfs -text
 Logo_Zymatica_Voice.png filter=lfs diff=lfs merge=lfs -text
+Zymatica_Voice_Lora_Guide.pdf filter=lfs diff=lfs merge=lfs -text

README.md

@@ -66,6 +66,7 @@ Each invention is isolated in its own folder and contains a complete academic **
 | **08** | [EPAUP Weight Projection](08_EPAUP_Weight_Projection) | Projects weights onto word embedding matrices. | [Whitepaper](08_EPAUP_Weight_Projection/WHITEPAPER.md) | [run_proof.py](08_EPAUP_Weight_Projection/run_proof.py) |
 | **09** | [Tokenizer Varint Coding](09_Tokenizer_Varint_Coding) | Prefix-suffix varint differential token coder. | [Whitepaper](09_Tokenizer_Varint_Coding/WHITEPAPER.md) | [run_proof.py](09_Tokenizer_Varint_Coding/run_proof.py) |
 | **10** | [Multi-Language Runtimes](10_Multi_Language_Runtimes) | Native runtimes (C++, Rust, Go, Swift, Java). | [Whitepaper](10_Multi_Language_Runtimes/WHITEPAPER.md) | [run_proof.py](10_Multi_Language_Runtimes/run_proof.py) |
+| **05/10** | [Zymatica Voice LoRa Guide](Zymatica_Voice_Lora_Guide.md) | AI Agent integration guide for physical LoRa hardware verification. | [MD Version](Zymatica_Voice_Lora_Guide.md) | [PDF Version](Zymatica_Voice_Lora_Guide.pdf) |
 | **11** | [RCRA Resonance Alignment](11_RCRA_Resonance_Alignment) | Fine-tuning using radical resonance loss. | [Whitepaper](11_RCRA_Resonance_Alignment/WHITEPAPER.md) | [run_proof.py](11_RCRA_Resonance_Alignment/run_proof.py) |
 | **12** | [Brand Assets Artwork](12_Brand_Assets_Artwork) | Official branding, logos, and design assets. | [Whitepaper](12_Brand_Assets_Artwork/WHITEPAPER.md) | [run_proof.py](12_Brand_Assets_Artwork/run_proof.py) |
 | **13** | [Multi-Centroid Steering](13_Multi_Centroid_Steering) | Dynamic English/CJK hidden state steering. | [Whitepaper](13_Multi_Centroid_Steering/WHITEPAPER.md) | [run_proof.py](13_Multi_Centroid_Steering/run_proof.py) |

Zymatica_Voice_Lora_Guide.pdf

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0dfb390f971cd0b154f0b22aec196d9ec1ae320a7924df0470975883bd902d3f
+size 917954

generate_voice_guide_pdf.py

@@ -0,0 +1,684 @@
+# -*- coding: utf-8 -*-
+import os
+import re
+import sys
+from reportlab.lib.pagesizes import letter
+from reportlab.platypus import SimpleDocTemplate, Paragraph, Spacer, Table, TableStyle, Image, PageBreak, KeepTogether
+from reportlab.lib.styles import getSampleStyleSheet, ParagraphStyle
+from reportlab.lib import colors
+from reportlab.pdfgen import canvas
+
+class NumberedCanvas(canvas.Canvas):
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self._saved_page_states = []
+
+    def showPage(self):
+        self._saved_page_states.append(dict(self.__dict__))
+        self._startPage()
+
+    def save(self):
+        num_pages = len(self._saved_page_states)
+        for state in self._saved_page_states:
+            self.__dict__.update(state)
+            self.draw_page_decorations(num_pages)
+            super().showPage()
+        super().save()
+
+    def draw_page_decorations(self, page_count):
+        self.saveState()
+        
+        # We start headers/footers on page 2
+        if self._pageNumber > 1:
+            # Running Header
+            self.setFont("Helvetica-Bold", 8)
+            self.setFillColor(colors.HexColor("#1A365D"))
+            self.drawString(54, 755, "ZYMATICA VOICE: A GUIDE TO LORA FOR AI AGENTS")
+            
+            self.setFont("Helvetica", 8)
+            self.setFillColor(colors.HexColor("#718096"))
+            self.drawRightString(558, 755, "IP CLASS 05/10 - TECHNICAL SPECIFICATION")
+            
+            # Header line
+            self.setStrokeColor(colors.HexColor("#CBD5E0"))
+            self.setLineWidth(0.75)
+            self.line(54, 747, 558, 747)
+            
+            # Running Footer
+            self.setStrokeColor(colors.HexColor("#CBD5E0"))
+            self.setLineWidth(0.75)
+            self.line(54, 55, 558, 55)
+            
+            self.setFont("Helvetica", 8)
+            self.setFillColor(colors.HexColor("#718096"))
+            self.drawString(54, 42, "© 2026 Zymatica.space | astronautshe.com | Devs One | We Are TheAiCollective.art")
+            
+            page_text = f"Page {self._pageNumber} of {page_count}"
+            self.drawRightString(558, 42, page_text)
+            
+        self.restoreState()
+
+def md_to_html(text):
+    # Escape '&' but avoid double escaping if it's already an entity
+    # A simple way is to replace '&' with '&' except for <, >, &, •, –, —
+    # Let's replace '&' first
+    text = text.replace("&", "&")
+    text = text.replace("&", "&")
+    text = text.replace("•", "•")
+    text = text.replace("–", "–")
+    text = text.replace("—", "—")
+    
+    # Replace '<' and '>' except when they look like HTML tags we want to support:
+    # <b>, </b>, <i>, </i>, <font ...>, </font>, <a>, </a>, <br/>, <br>
+    # We can temporarily hide our tags, clean the rest, and restore them, or just use regular expressions carefully.
+    
+    # Let's do markdown replacement
+    # Bold **text**
+    text = re.sub(r'\*\*(.*?)\*\*', r'<b>\1</b>', text)
+    # Italic *text*
+    text = re.sub(r'\*(.*?)\*', r'<i>\1</i>', text)
+    # Inline code `text`
+    text = re.sub(r'`(.*?)`', r'<font face="Courier" color="#2C5282">\1</font>', text)
+    # Links [text](url)
+    text = re.sub(r'\[(.*?)\]\((.*?)\)', r'<a href="\2"><font color="#2B6CB0"><u>\1</u></font></a>', text)
+    
+    return text
+
+def parse_markdown(filepath):
+    if not os.path.exists(filepath):
+        print(f"Error: {filepath} not found.")
+        sys.exit(1)
+        
+    with open(filepath, 'r', encoding='utf-8') as f:
+        lines = f.readlines()
+        
+    blocks = []
+    current_block = None
+    
+    in_code = False
+    code_lang = ""
+    code_lines = []
+    
+    in_table = False
+    table_lines = []
+    
+    in_quote = False
+    quote_lines = []
+    
+    for line_raw in lines:
+        line = line_raw.rstrip('\r\n')
+        line_stripped = line.strip()
+        
+        # Code block handler
+        if line_stripped.startswith('```'):
+            if in_code:
+                # End of code block
+                blocks.append({
+                    'type': 'code',
+                    'lang': code_lang,
+                    'content': '\n'.join(code_lines)
+                })
+                in_code = False
+                code_lines = []
+            else:
+                # Start of code block
+                in_code = True
+                code_lang = line_stripped[3:].strip()
+            continue
+            
+        if in_code:
+            code_lines.append(line)
+            continue
+            
+        # Table handler
+        if line_stripped.startswith('|'):
+            if not in_table:
+                in_table = True
+                table_lines = []
+            table_lines.append(line)
+            continue
+        elif in_table:
+            # Table ended
+            blocks.append({
+                'type': 'table',
+                'content': table_lines
+            })
+            in_table = False
+            table_lines = []
+            
+        # Blockquote handler
+        if line_stripped.startswith('>'):
+            if not in_quote:
+                in_quote = True
+                quote_lines = []
+            # Strip the leading '>' and space
+            content = line_stripped[1:].strip()
+            quote_lines.append(content)
+            continue
+        elif in_quote:
+            # Blockquote ended
+            blocks.append({
+                'type': 'quote',
+                'content': '\n'.join(quote_lines)
+            })
+            in_quote = False
+            quote_lines = []
+            
+        # Bullet list item handler
+        if line_stripped.startswith('* ') or line_stripped.startswith('- ') or re.match(r'^\d+\.\s', line_stripped):
+            is_ordered = bool(re.match(r'^\d+\.\s', line_stripped))
+            if is_ordered:
+                match = re.match(r'^(\d+)\.\s(.*)', line_stripped)
+                num = match.group(1)
+                text = match.group(2)
+                blocks.append({
+                    'type': 'list_item',
+                    'ordered': True,
+                    'number': num,
+                    'content': text
+                })
+            else:
+                text = line_stripped[2:]
+                blocks.append({
+                    'type': 'list_item',
+                    'ordered': False,
+                    'content': text
+                })
+            continue
+            
+        # Headers
+        if line_stripped.startswith('# '):
+            blocks.append({'type': 'h1', 'content': line_stripped[2:]})
+            continue
+        elif line_stripped.startswith('## '):
+            blocks.append({'type': 'h2', 'content': line_stripped[3:]})
+            continue
+        elif line_stripped.startswith('### '):
+            blocks.append({'type': 'h3', 'content': line_stripped[4:]})
+            continue
+            
+        # Horizontal rule
+        if line_stripped in ['---', '***']:
+            blocks.append({'type': 'hr'})
+            continue
+            
+        # Empty lines
+        if not line_stripped:
+            continue
+            
+        # Standard paragraph
+        blocks.append({'type': 'paragraph', 'content': line_stripped})
+        
+    # Flush remaining blocks
+    if in_code:
+        blocks.append({'type': 'code', 'lang': code_lang, 'content': '\n'.join(code_lines)})
+    if in_table:
+        blocks.append({'type': 'table', 'content': table_lines})
+    if in_quote:
+        blocks.append({'type': 'quote', 'content': '\n'.join(quote_lines)})
+        
+    return blocks
+
+def build_pdf(md_path, pdf_path):
+    print(f"Parsing markdown from: {md_path}")
+    blocks = parse_markdown(md_path)
+    
+    doc = SimpleDocTemplate(
+        pdf_path,
+        pagesize=letter,
+        leftMargin=54,
+        rightMargin=54,
+        topMargin=72,
+        bottomMargin=72
+    )
+    
+    styles = getSampleStyleSheet()
+    
+    # Custom Palette
+    primary_color = colors.HexColor("#1A365D")   # Deep Navy
+    secondary_color = colors.HexColor("#2B6CB0") # Slate Blue
+    dark_neutral = colors.HexColor("#2D3748")    # Charcoal
+    accent_color = colors.HexColor("#9B2C2C")    # Deep Crimson
+    light_bg = colors.HexColor("#F7FAFC")        # Warm White
+    border_color = colors.HexColor("#E2E8F0")    # Border Grey
+    
+    # Custom Styles
+    title_style = ParagraphStyle(
+        'DocTitle',
+        parent=styles['Heading1'],
+        fontName='Helvetica-Bold',
+        fontSize=20,
+        leading=24,
+        textColor=primary_color,
+        spaceAfter=4
+    )
+    
+    subtitle_style = ParagraphStyle(
+        'DocSubtitle',
+        parent=styles['Normal'],
+        fontName='Helvetica',
+        fontSize=11,
+        leading=15,
+        textColor=secondary_color,
+        spaceAfter=12
+    )
+    
+    meta_style = ParagraphStyle(
+        'DocMeta',
+        parent=styles['Normal'],
+        fontName='Helvetica-Bold',
+        fontSize=9.5,
+        leading=13,
+        textColor=dark_neutral,
+        spaceAfter=2
+    )
+    
+    h1_style = ParagraphStyle(
+        'SecHeading1',
+        parent=styles['Heading1'],
+        fontName='Helvetica-Bold',
+        fontSize=13.5,
+        leading=17,
+        textColor=primary_color,
+        spaceBefore=14,
+        spaceAfter=8,
+        keepWithNext=True
+    )
+    
+    h2_style = ParagraphStyle(
+        'SecHeading2',
+        parent=styles['Heading2'],
+        fontName='Helvetica-Bold',
+        fontSize=10.5,
+        leading=14,
+        textColor=secondary_color,
+        spaceBefore=10,
+        spaceAfter=6,
+        keepWithNext=True
+    )
+    
+    h3_style = ParagraphStyle(
+        'SecHeading3',
+        parent=styles['Heading3'],
+        fontName='Helvetica-Bold',
+        fontSize=9.5,
+        leading=13,
+        textColor=dark_neutral,
+        spaceBefore=8,
+        spaceAfter=4,
+        keepWithNext=True
+    )
+    
+    body_style = ParagraphStyle(
+        'BodyText',
+        parent=styles['Normal'],
+        fontName='Helvetica',
+        fontSize=9,
+        leading=13,
+        textColor=dark_neutral,
+        spaceAfter=6
+    )
+    
+    bullet_style = ParagraphStyle(
+        'BulletText',
+        parent=styles['Normal'],
+        fontName='Helvetica',
+        fontSize=8.5,
+        leading=12.5,
+        textColor=dark_neutral,
+        leftIndent=15,
+        firstLineIndent=-10,
+        spaceAfter=3
+    )
+    
+    code_style = ParagraphStyle(
+        'CodeText',
+        parent=styles['Normal'],
+        fontName='Courier',
+        fontSize=7.5,
+        leading=10,
+        textColor=colors.HexColor("#2C5282")
+    )
+    
+    quote_style = ParagraphStyle(
+        'QuoteText',
+        parent=styles['Normal'],
+        fontName='Helvetica-Oblique',
+        fontSize=8.5,
+        leading=12,
+        textColor=colors.HexColor("#2D3748")
+    )
+    
+    table_header_style = ParagraphStyle(
+        'TableHeader',
+        parent=styles['Normal'],
+        fontName='Helvetica-Bold',
+        fontSize=8,
+        leading=11,
+        textColor=colors.white
+    )
+    
+    table_cell_style = ParagraphStyle(
+        'TableCell',
+        parent=styles['Normal'],
+        fontName='Helvetica',
+        fontSize=7.5,
+        leading=10.5,
+        textColor=dark_neutral
+    )
+    
+    table_cell_bold = ParagraphStyle(
+        'TableCellBold',
+        parent=styles['Normal'],
+        fontName='Helvetica-Bold',
+        fontSize=7.5,
+        leading=10.5,
+        textColor=dark_neutral
+    )
+    
+    story = []
+    
+    # --- COVER PAGE ---
+    # Header branding table
+    logo_path = "Logo_Zymatica_Voice.png"
+    if not os.path.exists(logo_path):
+        # Fallback to J:\Language-U path
+        logo_path = "j:/Language-U/zymatica.space_repo/Logo_Zymatica_Voice.png"
+    if not os.path.exists(logo_path):
+        logo_path = "j:/Language-U/Logo_Zymatica_Voice.png"
+        
+    logo_exists = os.path.exists(logo_path)
+    
+    header_data = []
+    if logo_exists:
+        logo_img = Image(logo_path, width=54, height=54)
+        header_data = [[logo_img, Paragraph("<b>THE AI COLLECTIVE</b><br/><font color='#718096'>zymatica.space &bull; astronautshe.com &bull; Devs One</font>", subtitle_style)]]
+    else:
+        header_data = [[Paragraph("<b>THE AI COLLECTIVE</b><br/><font color='#718096'>zymatica.space &bull; astronautshe.com &bull; Devs One</font>", subtitle_style)]]
+        
+    header_table = Table(header_data, colWidths=[65, 439] if logo_exists else [504])
+    header_table.setStyle(TableStyle([
+        ('VALIGN', (0,0), (-1,-1), 'MIDDLE'),
+        ('BOTTOMPADDING', (0,0), (-1,-1), 10),
+    ]))
+    story.append(header_table)
+    story.append(Spacer(1, 10))
+    
+    # Separator Line
+    story.append(Table([[ "" ]], colWidths=[504], rowHeights=[2], style=TableStyle([
+        ('BACKGROUND', (0,0), (-1,-1), primary_color),
+        ('BOTTOMPADDING', (0,0), (-1,-1), 0),
+        ('TOPPADDING', (0,0), (-1,-1), 0),
+    ])))
+    story.append(Spacer(1, 15))
+    
+    # Document Title & Info
+    story.append(Paragraph("zymatica voice - A Guide To LoRa", title_style))
+    story.append(Paragraph("Zymatica Proprietary Protocol Specification | IP Class 05/10", subtitle_style))
+    
+    story.append(Paragraph("<b>Prepared For:</b> Autonomous AI Systems Agents, Coding Copilots, and Orchestrators", meta_style))
+    story.append(Paragraph("<b>Mission:</b> Execute, debug, and verify Language-U Semantic Communication loops over lossy physical LoRa channels.", meta_style))
+    story.append(Paragraph("<b>Authors & The AI Collective:</b> zymatica.space | astronautshe.com | Devs One (We Are TheAiCollective.art)", meta_style))
+    story.append(Spacer(1, 20))
+    
+    # Large Cover Logo Image if available
+    if logo_exists:
+        large_logo = Image(logo_path, width=160, height=160)
+        story.append(KeepTogether([
+            Spacer(1, 15),
+            Table([[large_logo]], colWidths=[504], style=TableStyle([
+                ('ALIGN', (0,0), (-1,-1), 'CENTER'),
+                ('BOTTOMPADDING', (0,0), (-1,-1), 15),
+            ])),
+            Spacer(1, 15),
+        ]))
+        
+    # Abstract Box or Callout Box at bottom of Page 1
+    notice_text = (
+        "<b>SYSTEM PROFILE SUMMARY:</b> This specification provides the complete hardware integration, "
+        "driver adjustments, validation parameters, and diagnostic routines for the RAKWireless RAK2287/RAK5146 gateway "
+        "concentrator modules based on the Semtech SX1302 LoRa baseband chip. It outlines reset sequences, I2C temperature sensor patches, "
+        "and dynamic verification scripts using LLD-AC range-coding and XOR-FEC packetization. Designed for direct parser parsing."
+    )
+    notice_table = Table([[ Paragraph(notice_text, table_cell_style) ]], colWidths=[504])
+    notice_table.setStyle(TableStyle([
+        ('BACKGROUND', (0,0), (-1,-1), colors.HexColor("#EDF2F7")),
+        ('BORDER', (0,0), (-1,-1), 0.75, colors.HexColor("#CBD5E0")),
+        ('PADDING', (0,0), (-1,-1), 10),
+    ]))
+    story.append(notice_table)
+    
+    story.append(PageBreak())
+    
+    # --- PARSING CONTENT ---
+    # We will build the remaining document sections
+    idx = 0
+    while idx < len(blocks):
+        block = blocks[idx]
+        b_type = block['type']
+        
+        if b_type == 'h1':
+            # We don't repeat the main page 1 title, but if it's there we can render it.
+            # Skip if it is the title since we did cover page
+            if "zymatica voice" in block['content'].lower():
+                idx += 1
+                continue
+            text = md_to_html(block['content'])
+            story.append(Paragraph(text, h1_style))
+            
+        elif b_type == 'h2':
+            # Skip branding headers already handled on cover
+            if "we are theaicollective.art" in block['content'].lower():
+                idx += 1
+                continue
+            text = md_to_html(block['content'])
+            story.append(Paragraph(text, h2_style))
+            
+        elif b_type == 'h3':
+            text = md_to_html(block['content'])
+            story.append(Paragraph(text, h3_style))
+            
+        elif b_type == 'paragraph':
+            # Skip licensing subheadings that belong to cover metadata
+            if "ip class 05/10" in block['content'].lower():
+                idx += 1
+                continue
+            text = md_to_html(block['content'])
+            story.append(Paragraph(text, body_style))
+            
+        elif b_type == 'list_item':
+            text = md_to_html(block['content'])
+            if block['ordered']:
+                bullet_prefix = f"<b>{block['number']}.</b> "
+                story.append(Paragraph(f"{bullet_prefix}{text}", bullet_style))
+            else:
+                bullet_prefix = "&bull; "
+                story.append(Paragraph(f"{bullet_prefix}{text}", bullet_style))
+                
+        elif b_type == 'code':
+            # Preformatted code blocks
+            code_content = block['content']
+            # Escape HTML characters so reportlab doesn't break
+            code_content = code_content.replace("&", "&amp;").replace("<", "&lt;").replace(">", "&gt;")
+            
+            # Format text into Paragraphs to support wrap-around (or pre-formatting style)
+            code_lines_flow = []
+            for c_line in code_content.splitlines():
+                # Retain indentation by replacing spaces with non-breaking spaces
+                c_line_indented = c_line.replace(" ", "&nbsp;")
+                code_lines_flow.append(Paragraph(c_line_indented, code_style))
+                
+            # Render code in a grey box Table
+            code_box_table = Table([[code_lines_flow]], colWidths=[504])
+            code_box_table.setStyle(TableStyle([
+                ('BACKGROUND', (0,0), (-1,-1), colors.HexColor("#F7FAFC")),
+                ('BORDER', (0,0), (-1,-1), 0.5, colors.HexColor("#CBD5E0")),
+                ('PADDING', (0,0), (-1,-1), 8),
+                ('TOPPADDING', (0,0), (-1,-1), 6),
+                ('BOTTOMPADDING', (0,0), (-1,-1), 6),
+            ]))
+            
+            story.append(KeepTogether([
+                Spacer(1, 4),
+                code_box_table,
+                Spacer(1, 6)
+            ]))
+            
+        elif b_type == 'quote':
+            quote_text = block['content']
+            
+            # Check if this is a caution box
+            is_caution = False
+            if "[!CAUTION]" in quote_text:
+                is_caution = True
+                quote_text = quote_text.replace("[!CAUTION]", "").strip()
+                
+            quote_html = md_to_html(quote_text)
+            quote_para = Paragraph(quote_html, quote_style)
+            
+            # Style the quote callout
+            if is_caution:
+                bg_col = colors.HexColor("#FFF5F5") # Reddish Alert
+                brd_col = colors.HexColor("#FEB2B2")
+                lbl_para = Paragraph("<b>⚠️ CAUTION: ANTENNA LOAD REQUIREMENT</b>", ParagraphStyle(
+                    'CautionLabel',
+                    parent=styles['Normal'],
+                    fontName='Helvetica-Bold',
+                    fontSize=8.5,
+                    leading=12,
+                    textColor=accent_color,
+                    spaceAfter=4
+                ))
+                quote_content_table = Table([[lbl_para], [quote_para]], colWidths=[490])
+            else:
+                bg_col = colors.HexColor("#EDF2F7") # Greyish Info
+                brd_col = colors.HexColor("#CBD5E0")
+                quote_content_table = Table([[quote_para]], colWidths=[490])
+                
+            quote_content_table.setStyle(TableStyle([
+                ('PADDING', (0,0), (-1,-1), 0),
+                ('VALIGN', (0,0), (-1,-1), 'TOP'),
+            ]))
+            
+            # Box wrapper with left accent border
+            quote_box = Table([[quote_content_table]], colWidths=[504])
+            quote_box.setStyle(TableStyle([
+                ('BACKGROUND', (0,0), (-1,-1), bg_col),
+                ('LINELEFT', (0,0), (0,-1), 4, accent_color if is_caution else secondary_color),
+                ('PADDING', (0,0), (-1,-1), 8),
+                ('TOPPADDING', (0,0), (-1,-1), 8),
+                ('BOTTOMPADDING', (0,0), (-1,-1), 8),
+                ('BORDER', (0,0), (-1,-1), 0.5, brd_col),
+            ]))
+            
+            story.append(KeepTogether([
+                Spacer(1, 6),
+                quote_box,
+                Spacer(1, 6)
+            ]))
+            
+        elif b_type == 'table':
+            # Parse MD table lines
+            table_lines = block['content']
+            
+            # Filter separator lines like |:---|---|
+            filtered_rows = []
+            for r_line in table_lines:
+                if re.match(r'^\|\s*[:\-]+\s*\|', r_line.strip()) or '---' in r_line:
+                    continue
+                filtered_rows.append(r_line)
+                
+            table_cells_data = []
+            for row_idx, r_line in enumerate(filtered_rows):
+                # Split cells, ignore first and last empty splits because of starting/ending |
+                cells = [c.strip() for c in r_line.split('|')]
+                if len(cells) > 1:
+                    # If line starts and ends with |, the split list has empty cells at boundaries
+                    if cells[0] == '':
+                        cells = cells[1:]
+                    if len(cells) > 0 and cells[-1] == '':
+                        cells = cells[:-1]
+                        
+                row_cells_flow = []
+                for cell in cells:
+                    cell_html = md_to_html(cell)
+                    if row_idx == 0:
+                        row_cells_flow.append(Paragraph(cell_html, table_header_style))
+                    else:
+                        # Decide if bold cell
+                        if cell.startswith('**') or cell.startswith('`'):
+                            row_cells_flow.append(Paragraph(cell_html, table_cell_bold))
+                        else:
+                            row_cells_flow.append(Paragraph(cell_html, table_cell_style))
+                if row_cells_flow:
+                    table_cells_data.append(row_cells_flow)
+                    
+            # Check number of columns to determine widths
+            if table_cells_data:
+                num_cols = len(table_cells_data[0])
+                # Distribute widths: 504 pt total
+                if num_cols == 3:
+                    # failure signature table: Error (110pt), Root Cause (120pt), Action (274pt)
+                    col_widths = [110, 120, 274]
+                else:
+                    col_widths = [504 / num_cols] * num_cols
+                    
+                md_table = Table(table_cells_data, colWidths=col_widths, repeatRows=1)
+                md_table.setStyle(TableStyle([
+                    ('BACKGROUND', (0,0), (-1,0), primary_color),
+                    ('ALIGN', (0,0), (-1,-1), 'LEFT'),
+                    ('VALIGN', (0,0), (-1,-1), 'TOP'),
+                    ('BOTTOMPADDING', (0,0), (-1,-1), 5),
+                    ('TOPPADDING', (0,0), (-1,-1), 5),
+                    ('LEFTPADDING', (0,0), (-1,-1), 5),
+                    ('RIGHTPADDING', (0,0), (-1,-1), 5),
+                    ('ROWBACKGROUNDS', (0,1), (-1,-1), [colors.white, colors.HexColor("#F7FAFC")]),
+                    ('GRID', (0,0), (-1,-1), 0.5, border_color),
+                ]))
+                
+                story.append(KeepTogether([
+                    Spacer(1, 6),
+                    md_table,
+                    Spacer(1, 6)
+                ]))
+                
+        elif b_type == 'hr':
+            story.append(Spacer(1, 8))
+            story.append(Table([[ "" ]], colWidths=[504], rowHeights=[1], style=TableStyle([
+                ('BACKGROUND', (0,0), (-1,-1), border_color),
+                ('BOTTOMPADDING', (0,0), (-1,-1), 0),
+                ('TOPPADDING', (0,0), (-1,-1), 0),
+            ])))
+            story.append(Spacer(1, 8))
+            
+        idx += 1
+        
+    # Signature block at the very end
+    story.append(Spacer(1, 15))
+    story.append(Table([[ "" ]], colWidths=[504], rowHeights=[1.5], style=TableStyle([
+        ('BACKGROUND', (0,0), (-1,-1), primary_color),
+        ('BOTTOMPADDING', (0,0), (-1,-1), 0),
+        ('TOPPADDING', (0,0), (-1,-1), 0),
+    ])))
+    story.append(Spacer(1, 10))
+    
+    sig_text = (
+        "<b>VERIFICATION SIGN OFF:</b><br/>"
+        "This specification is verified for execution by coding copilots and agent runtimes. "
+        "All parameters correspond to physical hardware EUI: <code>0x0016c001ff13ce58</code>.<br/>"
+        "<i>Gateway Integrator:</i> astronautshe.com &bull; "
+        "<i>Protocol Lead:</i> zymatica.space &bull; "
+        "<i>Orchestrator Agent:</i> Devs One &bull; "
+        "<i>Signed on behalf of:</i> TheAiCollective.art"
+    )
+    story.append(Paragraph(sig_text, body_style))
+    
+    print(f"Building PDF to: {pdf_path}")
+    doc.build(story, canvasmaker=NumberedCanvas)
+    print("[+] PDF built successfully.")
+
+if __name__ == "__main__":
+    base_dir = os.path.dirname(os.path.abspath(__file__))
+    md_file = os.path.join(base_dir, "Zymatica_Voice_Lora_Guide.md")
+    pdf_file = os.path.join(base_dir, "Zymatica_Voice_Lora_Guide.pdf")
+    build_pdf(md_file, pdf_file)