diff --git a/tiktok2bsky.py b/tiktok2bsky.py
index 66d6e8b..cb2d86f 100644
--- a/tiktok2bsky.py
+++ b/tiktok2bsky.py
@@ -789,40 +789,30 @@ def post_video_to_bluesky(
 # ─────────────────────────────────────────────────────────────────────────────
 #  Caption builder
 # ─────────────────────────────────────────────────────────────────────────────
-def build_caption(video_info: dict, tiktok_handle: str, max_len: int = 290) -> str:
+def build_caption(video_info: dict, tiktok_handle: str, max_len: int = 300) -> str:
     """
     Build a Bluesky post caption from video metadata.
 
-    Format:
-        <description>
-        <tiktok_url>
+    The TikTok URL is intentionally omitted — the video is already
+    embedded in the post, so the URL is redundant.
 
-    If description + URL exceeds 290 chars, the description is trimmed at
-    the last whitespace boundary before the limit to avoid cutting mid-word
-    or mid-hashtag.
+    If the description exceeds 300 chars it is trimmed at the last
+    whitespace boundary before the limit.
     """
     desc = (video_info.get("description") or "").strip()
-    url  = video_info.get("url", "").strip()
 
     if not desc:
-        # No caption available — just post the URL
-        return url
+        return ""
 
-    # Reserve space for newline + URL
-    url_block = f"\n{url}"
-    max_desc  = max_len - len(url_block)
-
-    if len(desc) > max_desc:
-        trimmed = desc[:max_desc - 1]
+    if len(desc) > max_len:
+        trimmed = desc[:max_len - 1]
         cut     = trimmed.rfind(" ")
         # Only use word boundary if it doesn't cut off too much
-        if cut > max_desc // 2:
+        if cut > max_len // 2:
             trimmed = trimmed[:cut]
         desc = trimmed + "…"
 
-    return f"{desc}{url_block}"
-
-
+    return desc
 # ─────────────────────────────────────────────────────────────────────────────
 #  TikTok scraping — Playwright
 # ─────────────────────────────────────────────────────────────────────────────