diff --git a/README.md b/README.md
index 74d01a8..b55c873 100644
--- a/README.md
+++ b/README.md
@@ -13,5 +13,14 @@ scrapling install
 ## Run
 
 ```bash
-python scrape_happyfappy_bookmarks.py --cookie-file cookies.txt --output bookmarks.json
+.venv/bin/python scrape_happyfappy_bookmarks.py --cookie-file cookies.txt --output bookmarks.json
+```
+
+## Download Single Torrent
+
+```bash
+.venv/bin/python download_happyfappy_torrent.py \
+  --url "https://www.happyfappy.net/torrents.php?id=110178" \
+  --cookie-file cookies.txt \
+  --output-dir torrent
 ```
diff --git a/download_happyfappy_torrent.py b/download_happyfappy_torrent.py
new file mode 100644
index 0000000..684caba
--- /dev/null
+++ b/download_happyfappy_torrent.py
@@ -0,0 +1,292 @@
+#!/usr/bin/env python3
+from __future__ import annotations
+
+import argparse
+import time
+from pathlib import Path
+from typing import Any
+from urllib.parse import urlparse
+
+from scrapling.fetchers import DynamicSession
+
+
+def _domain_matches(target_host: str, cookie_domain: str) -> bool:
+    cd = cookie_domain.lstrip(".").lower()
+    th = target_host.lower()
+    return th == cd or th.endswith("." + cd)
+
+
+def parse_cookie_string(cookie_string: str, target_host: str) -> dict[str, str]:
+    cookies: dict[str, str] = {}
+    lines = cookie_string.splitlines()
+    looks_like_netscape = len(lines) > 1 and any("\t" in line for line in lines)
+
+    if looks_like_netscape:
+        for raw_line in lines:
+            line = raw_line.strip()
+            if not line or line.startswith("#"):
+                continue
+            parts = line.split("\t")
+            if len(parts) < 7:
+                continue
+            domain, _flag, _path, _secure, _expires, name, value = parts[:7]
+            if not _domain_matches(target_host, domain):
+                continue
+            if name:
+                cookies[name] = value
+        return cookies
+
+    for chunk in cookie_string.split(";"):
+        piece = chunk.strip()
+        if not piece or "=" not in piece:
+            continue
+        key, value = piece.split("=", 1)
+        key = key.strip()
+        value = value.strip()
+        if key:
+            cookies[key] = value
+    return cookies
+
+
+def parse_cookies_for_playwright(
+    cookie_string: str, target_host: str, base_url: str
+) -> list[dict[str, Any]]:
+    lines = cookie_string.splitlines()
+    cookies: list[dict[str, Any]] = []
+    looks_like_netscape = len(lines) > 1 and any("\t" in line for line in lines)
+
+    if looks_like_netscape:
+        for raw_line in lines:
+            line = raw_line.strip()
+            if not line or line.startswith("#"):
+                continue
+            parts = line.split("\t")
+            if len(parts) < 7:
+                continue
+            domain, _flag, path, secure, expires, name, value = parts[:7]
+            if not _domain_matches(target_host, domain):
+                continue
+            if not name:
+                continue
+
+            cookie_obj: dict[str, Any] = {
+                "name": name,
+                "value": value,
+                "domain": domain.lstrip("."),
+                "path": path or "/",
+                "secure": (secure.upper() == "TRUE"),
+            }
+            if expires.isdigit():
+                exp_num = int(expires)
+                if exp_num > 0:
+                    cookie_obj["expires"] = float(exp_num)
+            cookies.append(cookie_obj)
+        return cookies
+
+    kv = parse_cookie_string(cookie_string, target_host)
+    for name, value in kv.items():
+        cookies.append({"name": name, "value": value, "url": base_url})
+    return cookies
+
+
+def absolute_url(base_url: str, href: str) -> str:
+    href = href.strip()
+    if href.startswith("http://") or href.startswith("https://"):
+        return href
+    if href.startswith("/"):
+        return f"{base_url.rstrip('/')}{href}"
+    return f"{base_url.rstrip('/')}/{href}"
+
+
+def fetch_dynamic_with_retry(session: Any, url: str, retries: int, backoff_base: float) -> Any:
+    last_error: Exception | None = None
+    for attempt in range(retries):
+        try:
+            response = session.fetch(
+                url,
+                timeout=45_000,
+                load_dom=True,
+                network_idle=False,
+                google_search=False,
+            )
+            status = response.status
+            if status in (403, 404, 429) or status >= 500:
+                raise RuntimeError(f"HTTP {status}")
+            return response
+        except Exception as err:  # noqa: BLE001
+            last_error = err
+            if attempt == retries - 1:
+                break
+            time.sleep(backoff_base * (2**attempt))
+    raise RuntimeError(f"Request failed for {url}: {last_error}") from last_error
+
+
+def download_via_browser_with_retry(
+    session: DynamicSession, detail_url: str, retries: int, backoff_base: float
+) -> tuple[str, bytes]:
+    last_error: Exception | None = None
+    for attempt in range(retries):
+        page = session.context.new_page()
+        try:
+            page.goto(detail_url, wait_until="domcontentloaded", timeout=45_000)
+            with page.expect_download(timeout=45_000) as download_info:
+                clicked = False
+                selectors = [
+                    'span.torrent_buttons a[title*="Download"][href*="action=download"]',
+                    'a.button.blueButton[href*="action=download"]',
+                    'a[href*="action=download"][href*="torrent_pass"]',
+                ]
+                for selector in selectors:
+                    locator = page.locator(selector)
+                    if locator.count() > 0:
+                        locator.first.click()
+                        clicked = True
+                        break
+
+                if not clicked:
+                    locator = page.locator(
+                        "xpath=//a[contains(translate(normalize-space(string(.)),'abcdefghijklmnopqrstuvwxyz','ABCDEFGHIJKLMNOPQRSTUVWXYZ'),'DOWNLOAD') and contains(@href,'action=download')]"
+                    )
+                    if locator.count() > 0:
+                        locator.first.click()
+                        clicked = True
+
+                if not clicked:
+                    raise RuntimeError("Download button not found in interactive page.")
+
+            download = download_info.value
+            temp_path = download.path()
+            if not temp_path:
+                raise RuntimeError("Downloaded file path is empty.")
+            data = Path(temp_path).read_bytes()
+            filename = (download.suggested_filename or "downloaded.torrent").strip()
+            if not filename:
+                filename = "downloaded.torrent"
+            return filename, data
+        except Exception as err:  # noqa: BLE001
+            last_error = err
+            if attempt == retries - 1:
+                break
+            time.sleep(backoff_base * (2**attempt))
+        finally:
+            page.close()
+    raise RuntimeError(f"Request failed for {detail_url}: {last_error}") from last_error
+
+
+def find_download_link(response: Any) -> str:
+    selectors = [
+        'span.torrent_buttons a[title*="Download"][href*="action=download"]::attr(href)',
+        'a.button.blueButton[href*="action=download"]::attr(href)',
+        'a[href*="action=download"][href*="torrent_pass"]::attr(href)',
+    ]
+    for sel in selectors:
+        href = (response.css(sel).get("") or "").strip()
+        if href:
+            return href
+
+    # Fallback using text match if classes/attributes drift
+    href = (
+        response.xpath(
+            "//a[contains(translate(normalize-space(string(.)),"
+            "'abcdefghijklmnopqrstuvwxyz','ABCDEFGHIJKLMNOPQRSTUVWXYZ'),'DOWNLOAD') and contains(@href,'action=download')]/@href"
+        ).get("")
+        or ""
+    ).strip()
+    return href
+
+
+def normalize_filename(filename: str, download_url: str) -> str:
+    safe_name = Path(filename).name.strip()
+    if safe_name:
+        return safe_name if safe_name.lower().endswith(".torrent") else f"{safe_name}.torrent"
+    from_url = Path(urlparse(download_url).path).name.strip()
+    if from_url:
+        return from_url if from_url.lower().endswith(".torrent") else f"{from_url}.torrent"
+    return "downloaded.torrent"
+
+
+def looks_like_torrent_bytes(data: bytes) -> bool:
+    # Basic bencode sanity check for torrent files
+    return bool(data) and data.startswith(b"d") and (b"4:info" in data[:4096])
+
+
+def validate_torrent_response(download_url: str, filename: str, data: bytes) -> None:
+    good_ext = filename.lower().endswith(".torrent") or urlparse(download_url).path.lower().endswith(".torrent")
+    if not good_ext:
+        raise RuntimeError("Downloaded content has no .torrent extension in URL/filename.")
+    if not looks_like_torrent_bytes(data):
+        raise RuntimeError("Downloaded file failed torrent bencode check.")
+
+
+def run(args: argparse.Namespace) -> None:
+    base_url = args.base_url.rstrip("/")
+    target_host = urlparse(base_url).hostname or "www.happyfappy.net"
+
+    cookie_value = args.cookie or ""
+    if not cookie_value and args.cookie_file:
+        cookie_value = Path(args.cookie_file).read_text(encoding="utf-8").strip()
+    if not cookie_value:
+        raise ValueError("Cookie is required. Use --cookie or --cookie-file.")
+
+    cookies = parse_cookie_string(cookie_value, target_host=target_host)
+    if not cookies:
+        raise ValueError("No valid cookies parsed for target host.")
+    pw_cookies = parse_cookies_for_playwright(cookie_value, target_host=target_host, base_url=base_url)
+    if not pw_cookies:
+        raise ValueError("No Playwright-compatible cookies generated for target host.")
+
+    output_dir = Path(args.output_dir).resolve()
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+    with DynamicSession(
+        headless=True,
+        disable_resources=True,
+        cookies=pw_cookies,
+        google_search=False,
+        retries=1,
+        retry_delay=1,
+    ) as session:
+        detail_response = fetch_dynamic_with_retry(
+            session, args.url, retries=args.retries, backoff_base=args.backoff_base
+        )
+        href = find_download_link(detail_response)
+        if not href:
+            raise RuntimeError("Download link not found on page.")
+
+        download_url = absolute_url(base_url, href)
+        suggested_filename, data = download_via_browser_with_retry(
+            session, args.url, retries=args.retries, backoff_base=args.backoff_base
+        )
+        filename = normalize_filename(suggested_filename, download_url)
+        validate_torrent_response(download_url, filename, data)
+        output_path = output_dir / filename
+        output_path.write_bytes(data)  # overwrite behavior by design
+        print(f"Saved torrent to {output_path}")
+
+
+def make_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(
+        description="Download a torrent file from a single HappyFappy torrent detail page URL.",
+    )
+    parser.add_argument("--url", required=True, help="Torrent detail page URL")
+    parser.add_argument("--base-url", default="https://www.happyfappy.net")
+    parser.add_argument("--cookie", help='Raw cookie string, e.g. "a=1; b=2"')
+    parser.add_argument("--cookie-file", help="Path to cookie file")
+    parser.add_argument("--output-dir", default="torrent")
+    parser.add_argument("--retries", type=int, default=3)
+    parser.add_argument("--backoff-base", type=float, default=5.0)
+    return parser
+
+
+def main() -> None:
+    parser = make_parser()
+    args = parser.parse_args()
+    if args.retries < 1:
+        raise ValueError("--retries must be at least 1.")
+    if args.backoff_base < 0:
+        raise ValueError("--backoff-base must be >= 0.")
+    run(args)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/scrape_happyfappy_bookmarks.py b/scrape_happyfappy_bookmarks.py
index 5ac9f06..b79e549 100644
--- a/scrape_happyfappy_bookmarks.py
+++ b/scrape_happyfappy_bookmarks.py
@@ -5,21 +5,12 @@ import argparse
 import json
 import random
 import re
-import sys
 import time
 from pathlib import Path
 from typing import Any
 from urllib.parse import urlparse
 
-try:
-    from scrapling.fetchers import DynamicSession
-except ModuleNotFoundError:
-    local_repo = Path(__file__).resolve().parent / "Scrapling"
-    if local_repo.exists():
-        sys.path.insert(0, str(local_repo))
-        from scrapling.fetchers import DynamicSession
-    else:
-        raise
+from scrapling.fetchers import DynamicSession
 
 STOP_TEXT = "You have not bookmarked any torrents."
 BG_URL_RE = re.compile(r"url\((?:'|\")?(.*?)(?:'|\")?\)")
@@ -122,11 +113,13 @@ def extract_background_image(style: str) -> str | None:
     return value or None
 
 
-def extract_torrent_cards(response: Any) -> list[dict[str, Any]]:
+def extract_torrent_cards(response: Any, base_url: str) -> list[dict[str, Any]]:
     records: list[dict[str, Any]] = []
     cards = response.css("div.torrent_grid div.torrent_grid__torrent")
     for card in cards:
         page_url = (card.css('a[href^="/torrents.php?id="]::attr(href)').get("") or "").strip()
+        if page_url and not page_url.startswith("http"):
+            page_url = f"{base_url.rstrip('/')}{page_url}"
         category = (card.css("span.torrent_grid__torrent__cat::text").get("") or "").strip()
         title = (
             card.css("h3.trim::attr(title)").get("")
@@ -220,7 +213,7 @@ def run(args: argparse.Namespace) -> None:
             if should_stop(response):
                 break
 
-            page_records = extract_torrent_cards(response)
+            page_records = extract_torrent_cards(response, args.base_url)
             all_records.extend(page_records)
             print(f"[page={page}] extracted={len(page_records)} total={len(all_records)}")
             page += 1