diff --git a/fetch_blog.py b/fetch_blog.py
new file mode 100644
index 0000000..d9a7837
--- /dev/null
+++ b/fetch_blog.py
@@ -0,0 +1,43 @@
+import csv
+import os
+import requests
+from markdownify import markdownify as md
+
+CSV_PATH = "C:/projects/forth/bootslop/references/TimothyLottesBlog.csv"
+OUT_DIR = "C:/projects/forth/bootslop/references/TimothyLottesBlog"
+
+os.makedirs(OUT_DIR, exist_ok=True)
+
+with open(CSV_PATH, "r", encoding="utf-8") as f:
+    reader = csv.reader(f)
+    for row in reader:
+        if len(row) < 2:
+            continue
+        title = row[0].strip()
+        url = row[1].strip()
+        
+        # Clean up title for filename
+        filename = "".join(c for c in title if c.isalnum() or c in (' ', '-', '_')).strip()
+        filename = filename.replace(' ', '_') + ".md"
+        out_path = os.path.join(OUT_DIR, filename)
+        
+        if os.path.exists(out_path):
+            print(f"Already exists: {filename}")
+            continue
+            
+        print(f"Fetching: {title}")
+        try:
+            response = requests.get(url)
+            response.raise_for_status()
+            
+            # Convert HTML to markdown
+            markdown_text = md(response.text, heading_style="ATX")
+            
+            with open(out_path, "w", encoding="utf-8") as out_f:
+                out_f.write(f"# {title}\n\n")
+                out_f.write(f"**Source:** {url}\n\n")
+                out_f.write(markdown_text)
+                
+            print(f"Saved: {filename}")
+        except Exception as e:
+            print(f"Failed to fetch {title}: {e}")
diff --git a/fetch_notes.py b/fetch_notes.py
new file mode 100644
index 0000000..6a52e75
--- /dev/null
+++ b/fetch_notes.py
@@ -0,0 +1,76 @@
+import csv
+import os
+import requests
+from markdownify import markdownify as md
+
+NOTES_CSV = "C:/projects/forth/bootslop/references/FORTH_NOTES.csv"
+BLOG_CSV = "C:/projects/forth/bootslop/references/TimothyLottesBlog.csv"
+OUT_DIR = "C:/projects/forth/bootslop/references/ForthNotes"
+
+os.makedirs(OUT_DIR, exist_ok=True)
+
+# Load URLs we already processed in TimothyLottesBlog.csv
+ignored_urls = set()
+if os.path.exists(BLOG_CSV):
+    with open(BLOG_CSV, "r", encoding="utf-8") as f:
+        reader = csv.reader(f)
+        for row in reader:
+            if len(row) >= 2:
+                ignored_urls.add(row[1].strip())
+
+def is_relevant(title, url):
+    # Ignore YouTube as it's not useful to scrape HTML for them
+    if "youtube.com" in url or "youtu.be" in url:
+        return False
+    # Ignore PDFs (requires special handling)
+    if url.endswith(".pdf"):
+        return False
+    # Ignore twitter (requires JS/auth)
+    if "x.com" in url or "twitter.com" in url:
+        return False
+    return True
+
+with open(NOTES_CSV, "r", encoding="utf-8") as f:
+    reader = csv.reader(f)
+    next(reader, None) # Skip header
+    for row in reader:
+        if len(row) < 2:
+            continue
+            
+        title = row[0].strip()
+        url = row[1].strip()
+        
+        if url in ignored_urls:
+            continue
+            
+        if not is_relevant(title, url):
+            print(f"Skipping unsupported/media link: {title} ({url})")
+            continue
+            
+        # Clean up title for filename
+        filename = "".join(c for c in title if c.isalnum() or c in (' ', '-', '_')).strip()
+        filename = filename.replace(' ', '_') + ".md"
+        out_path = os.path.join(OUT_DIR, filename)
+        
+        if os.path.exists(out_path):
+            print(f"Already exists: {filename}")
+            continue
+            
+        print(f"Fetching: {title} from {url}")
+        try:
+            # Set a user-agent to avoid simple anti-bot blocks
+            headers = {'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'}
+            response = requests.get(url, headers=headers, timeout=10)
+            response.raise_for_status()
+            
+            # Convert HTML to markdown
+            markdown_text = md(response.text, heading_style="ATX")
+            
+            with open(out_path, "w", encoding="utf-8") as out_f:
+                out_f.write(f"# {title}\n\n")
+                out_f.write(f"**Source:** {url}\n\n")
+                out_f.write(markdown_text)
+                
+            print(f"Saved: {filename}")
+        except Exception as e:
+            print(f"Failed to fetch {title}: {e}")
diff --git a/image.png b/image.png
new file mode 100644
index 0000000..2646efc
Binary files /dev/null and b/image.png differ
diff --git a/ocr_interaction.py b/ocr_interaction.py
new file mode 100644
index 0000000..0e45ae9
--- /dev/null
+++ b/ocr_interaction.py
@@ -0,0 +1,39 @@
+import asyncio
+import os
+from winsdk.windows.storage import StorageFile
+from winsdk.windows.graphics.imaging import BitmapDecoder
+from winsdk.windows.media.ocr import OcrEngine
+from winsdk.windows.globalization import Language
+
+async def ocr_single_image(img_path):
+    if not os.path.exists(img_path):
+        print(f"File not found: {img_path}")
+        return
+
+    file = await StorageFile.get_file_from_path_async(os.path.abspath(img_path))
+    stream = await file.open_read_async()
+    decoder = await BitmapDecoder.create_async(stream)
+    bitmap = await decoder.get_software_bitmap_async()
+    
+    engine = OcrEngine.try_create_from_language(Language("en-US"))
+    if not engine:
+        print("Failed to create OCR engine")
+        return
+        
+    result = await engine.recognize_async(bitmap)
+    
+    output = [f"# OCR Thread: {os.path.basename(img_path)}\n"]
+    if result and result.lines:
+        for line in result.lines:
+            output.append(line.text)
+            
+    out_path = img_path + ".ocr.md"
+    # Using join with a literal newline to avoid potential issues in some environments
+    content = "\n".join(output)
+    with open(out_path, "w", encoding="utf-8") as f:
+        f.write(content)
+    print(f"OCR result saved to {out_path}")
+
+if __name__ == "__main__":
+    img = r"C:\projects\forth\bootslop\references\X.com - Onat & Lottes Interaction 1.png"
+    asyncio.run(ocr_single_image(img))
diff --git a/process_visuals.py b/process_visuals.py
index 760d236..705da6e 100644
--- a/process_visuals.py
+++ b/process_visuals.py
@@ -161,7 +161,7 @@ async def process_video(video_path, timestamps, video_name):
             
             code_crop = frame[min_y:max_y, min_x:max_x]
             code_path = os.path.join(out_dir, f"code_{t_sec:04d}s.jpg")
-            cv2.imwrite(code_path, code_crop)
+            Image.fromarray(cv2.cvtColor(code_crop, cv2.COLOR_BGR2RGB)).save(code_path)
             markdown_lines.append(f"\n*Saved code image: {code_path}*")
             
             # Find non-text visual content (e.g. diagrams) outside the text area
@@ -183,7 +183,7 @@ async def process_video(video_path, timestamps, video_name):
                     if bw > 100 and bh > 100: # large enough
                         visual_crop = frame[max(0, by-pad):min(frame.shape[0], by+bh+pad), max(0, bx-pad):min(frame.shape[1], bx+bw+pad)]
                         visual_path = os.path.join(out_dir, f"visual_{t_sec:04d}s.jpg")
-                        cv2.imwrite(visual_path, visual_crop)
+                        Image.fromarray(cv2.cvtColor(visual_crop, cv2.COLOR_BGR2RGB)).save(visual_path)
                         markdown_lines.append(f"\n*Saved non-text visual: {visual_path}*")
                         
         markdown_lines.append("\n---\n")