diff --git a/nfuck/__init__.py b/nfuck/__init__.py
index dec5438..40987cc 100644
--- a/nfuck/__init__.py
+++ b/nfuck/__init__.py
@@ -15,7 +15,8 @@ from nfuck.utils import sanitize_link
 
 dp = Dispatcher()
 
-SILENT_REMOVAL_IDS: set[int] = set(list(map(int, getenv("SILENT_REMOVAL_IDS", "").split(","))))
+SILENT_REMOVAL_IDS: set[int] = set(list(map(int, filter(lambda v: v, getenv("SILENT_REMOVAL_IDS", "").split(",")))))
+
 
 @dp.message(Command("check"))
 async def on_check(message: Message):
@@ -66,8 +67,12 @@ async def on_message(message: Message):
             if confidence > 0.9:
                 detected_links.append((entity.url, confidence))
     if detected_links:
+        await message.delete()
         if message.from_user and message.chat.id not in SILENT_REMOVAL_IDS:
-            msg = await message.reply(
+            if not message.bot:
+                raise RuntimeError("what")
+            msg = await message.bot.send_message(
+                message.chat.id,
                 str.join(
                     "\n",
                     [
@@ -82,13 +87,11 @@ async def on_message(message: Message):
                             ],
                         ),
                         f"Sender: {message.from_user.full_name} #{message.from_user.id} (@{message.from_user.username})",
-                        "(message will be deleted in 10 seconds)"
+                        "(message will be deleted in 10 seconds)",
+                        "False positive? Report <a href=\"https://forms.gle/cwj565M3y928M47g7\">here</a>!"
                     ],
                 ),
                 parse_mode="html",
             )
-            await message.delete()
             await sleep(10)
             await msg.delete()
-        else:
-            await message.delete()
diff --git a/nfuck/link_verifier.py b/nfuck/link_verifier.py
index f9ea4e3..eb995df 100644
--- a/nfuck/link_verifier.py
+++ b/nfuck/link_verifier.py
@@ -16,6 +16,10 @@ USER_AGENT = [
     "Mozilla/5.0 (X11; Linux x86_64; rv:122.0) Gecko/20100101 Firefox/122.0"
 ]
 
+URL_PATTERNS: list[tuple[float, Pattern, str]] = [
+    (10.0, regexp(r"https://t.me/\w+[bB]ot/claim"), "Telegram Bot claim link")
+]
+
 REGEX_PATTERNS: list[tuple[float, Pattern, str]] = [
     (1.0, regexp(r"\bp2e\b", IGNORECASE), "Play-to-earn keyword"),
     (5.0, regexp(r"play\-to\-earn", IGNORECASE), "Play-to-earn directly"),
@@ -32,7 +36,8 @@ REGEX_PATTERNS: list[tuple[float, Pattern, str]] = [
     (3.0, regexp(r"A collection of \w+ NFTs", IGNORECASE), "Collection of [some] NFTs"),
 ]
 
-MAX_SCORE = sum(t[0] for t in REGEX_PATTERNS)
+MAX_REGEX_SCORE = 30 # sum(t[0] for t in REGEX_PATTERNS)
+MAX_URL_SCORE = 10
 
 
 def explain_verification(content: str) -> list[tuple[float, str, Match]]:
@@ -56,6 +61,11 @@ async def verify_link(url: str) -> float:
     if any(fnmatch(domain, pat) for pat in DOMAIN_WHITELIST):
         logger.info("Score for %r: 0 (whitelisted domain)", url)
         return 0
+    for score, regex, explanation in REGEX_PATTERNS:
+        for match in regex.finditer(url):
+            total_score += score
+    if total_score >= MAX_REGEX_SCORE:
+        return total_score / MAX_REGEX_SCORE
     async with AsyncClient(
         headers={"User-Agent": get_random_useragent()}
     ) as client:
@@ -64,4 +74,4 @@ async def verify_link(url: str) -> float:
             logger.debug("%s: %s at %d", url, explanation, match.start())
             total_score += score
     logger.info("Score for %r: %f", url, total_score)
-    return total_score / MAX_SCORE
+    return total_score / MAX_REGEX_SCORE