diff --git a/.gitignore b/.gitignore
index 68bc17f..3b072b5 100644
--- a/.gitignore
+++ b/.gitignore
@@ -157,4 +157,7 @@ cython_debug/
 #  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
-#.idea/
+.idea/
+data/
+bot*
+defs/glover.py
diff --git a/defs/feed.py b/defs/feed.py
new file mode 100644
index 0000000..38c2116
--- /dev/null
+++ b/defs/feed.py
@@ -0,0 +1,64 @@
+import traceback
+from datetime import datetime
+from typing import List, Optional
+
+from bs4 import BeautifulSoup
+
+from init import request
+from defs.glover import rss_hub_host
+from defs.models import Tweet, User
+from feedparser import parse, FeedParserDict
+
+
+class UsernameNotFound(Exception):
+    pass
+
+
+async def get(username: str):
+    url = f"{rss_hub_host}/twitter/user/{username}"
+    response = await request.get(url)
+    if response.status_code == 200:
+        return parse(response.text)
+    elif response.status_code == 404:
+        raise UsernameNotFound
+    else:
+        return None
+
+
+async def parse_tweets(data: List[FeedParserDict]) -> List[Tweet]:
+    tweets = []
+    for tweet in data:
+        try:
+            description = tweet.get("description", "")
+            soup = BeautifulSoup(description, "lxml")
+            content = soup.get_text()
+            img_tag = soup.find_all("img")
+            images = [img.get("src") for img in img_tag if img.get("src")]
+            url = tweet.get("link", "")
+            time = datetime.strptime(tweet.get("published", ""), "%a, %d %b %Y %H:%M:%S %Z")
+            tweets.append(
+                Tweet(
+                    content=content,
+                    url=url,
+                    time=time,
+                    images=images
+                )
+            )
+        except Exception:
+            traceback.print_exc()
+    return tweets
+
+
+async def parse_user(username: str, data: FeedParserDict) -> User:
+    title = data.get("feed", {}).get("title", "")
+    name = title.replace("Twitter @", "")
+    tweets = await parse_tweets(data.get("entries", []))
+    return User(username=username, name=name, tweets=tweets)
+
+
+async def get_user(username: str) -> Optional[User]:
+    data = await get(username)
+    if data:
+        return await parse_user(username, data)
+    else:
+        return None
diff --git a/defs/models.py b/defs/models.py
new file mode 100644
index 0000000..15329bc
--- /dev/null
+++ b/defs/models.py
@@ -0,0 +1,33 @@
+from datetime import datetime
+from typing import List
+
+from pydantic import BaseModel
+
+
+class Tweet(BaseModel):
+    content: str
+    url: str
+    time: datetime
+    images: List[str]
+
+    @property
+    def id(self) -> int:
+        return int(self.url.split("/")[-1])
+
+    @property
+    def time_str(self) -> str:
+        return self.time.strftime("%Y-%m-%d %H:%M:%S")
+
+
+class User(BaseModel):
+    username: str
+    name: str
+    tweets: List[Tweet]
+
+    @property
+    def link(self) -> str:
+        return f"https://twitter.com/{self.username}"
+
+    @property
+    def format(self) -> str:
+        return f'<a href="{self.link}">{self.name}</a>'
diff --git a/defs/sqlite.py b/defs/sqlite.py
new file mode 100644
index 0000000..13855b5
--- /dev/null
+++ b/defs/sqlite.py
@@ -0,0 +1,49 @@
+from pathlib import Path
+from typing import List
+
+from sqlitedict import SqliteDict
+
+data_path = Path("data")
+data_path.mkdir(exist_ok=True)
+db_path = data_path / "data.db"
+db = SqliteDict(db_path, autocommit=True)
+
+
+class TweetDB:
+    prefix = "tweet_"
+
+    @staticmethod
+    def get_all(username: str) -> List[int]:
+        return db.get(f"{TweetDB.prefix}{username}", [])
+
+    @staticmethod
+    def check_id(username: str, tid: int) -> bool:
+        return tid in TweetDB.get_all(username)
+
+    @staticmethod
+    def add(username: str, tid: int) -> None:
+        tweets = TweetDB.get_all(username)
+        tweets.append(tid)
+        db[f"{TweetDB.prefix}{username}"] = tweets
+
+
+class UserDB:
+    @staticmethod
+    def get_all() -> List[str]:
+        return db.get("users", [])
+
+    @staticmethod
+    def check(username: str) -> bool:
+        return username in UserDB.get_all()
+
+    @staticmethod
+    def add(username: str) -> None:
+        users = UserDB.get_all()
+        users.append(username)
+        db["users"] = users
+
+    @staticmethod
+    def remove(username: str) -> None:
+        users = UserDB.get_all()
+        users.remove(username)
+        db["users"] = users
diff --git a/defs/update.py b/defs/update.py
new file mode 100644
index 0000000..a98be2a
--- /dev/null
+++ b/defs/update.py
@@ -0,0 +1,130 @@
+import asyncio
+import traceback
+from typing import List
+
+from pyrogram.enums import ParseMode
+from pyrogram.errors import FloodWait
+from pyrogram.types import InlineKeyboardMarkup, InlineKeyboardButton, InputMediaPhoto
+
+from defs.glover import cid, tid, owner
+from defs.models import User, Tweet
+from init import bot, logs
+from defs.sqlite import TweetDB, UserDB
+from defs.feed import get_user, UsernameNotFound
+
+
+def get_button(user: User, tweet: Tweet) -> InlineKeyboardMarkup:
+    return InlineKeyboardMarkup(
+        [
+            [
+                InlineKeyboardButton("Source", url=tweet.url),
+                InlineKeyboardButton("Author", url=user.link),
+            ]
+        ]
+    )
+
+
+def get_media_group(text: str, tweet: Tweet) -> List[InputMediaPhoto]:
+    data = []
+    images = tweet.images[:10]
+    for idx, image in enumerate(images):
+        data.append(
+            InputMediaPhoto(
+                image,
+                caption=text if idx == 0 else None,
+                parse_mode=ParseMode.HTML,
+            )
+        )
+    return data
+
+
+def flood_wait():
+    def decorator(function):
+        async def wrapper(*args, **kwargs):
+            try:
+                return await function(*args, **kwargs)
+            except FloodWait as e:
+                logs.warning(f"遇到 FloodWait，等待 {e.value} 秒后重试！")
+                await asyncio.sleep(e.value + 1)
+                return await wrapper(*args, **kwargs)
+            except Exception as e:
+                traceback.format_exc()
+                raise e
+
+        return wrapper
+
+    return decorator
+
+
+@flood_wait()
+async def send_to_user(user: User, tweet: Tweet):
+    text = "<b>Twitter Timeline Update</b>\n\n<code>"
+    text += tweet.content
+    text += f"</code>\n\n{user.format} 发表于 {tweet.time_str}"
+    if not tweet.images:
+        return await bot.send_message(
+            cid,
+            text,
+            disable_web_page_preview=True,
+            reply_to_message_id=tid,
+            parse_mode=ParseMode.HTML,
+            reply_markup=get_button(user, tweet),
+        )
+    elif len(tweet.images) == 1:
+        return await bot.send_photo(
+            cid,
+            tweet.images[0],
+            caption=text,
+            reply_to_message_id=tid,
+            parse_mode=ParseMode.HTML,
+            reply_markup=get_button(user, tweet),
+        )
+    else:
+        await bot.send_media_group(
+            cid,
+            get_media_group(text, tweet),
+            reply_to_message_id=tid,
+        )
+
+
+@flood_wait()
+async def send_username_changed(user: str):
+    text = f"获取 {user} 的数据失败，可能用户名已改变"
+    await bot.send_message(owner, text)
+
+
+async def send_check(user_data: User):
+    need_send_tweets = [
+        tweet for tweet in user_data.tweets
+        if not TweetDB.check_id(user_data.username, tweet.id)
+    ]
+    logs.info(f"需要推送 {len(need_send_tweets)} 条推文")
+    for tweet in need_send_tweets:
+        try:
+            await send_to_user(user_data, tweet)
+        except Exception:
+            logs.error(f"推送 {user_data.name} 的推文 {tweet.id} 失败")
+            traceback.print_exc()
+        TweetDB.add(user_data.username, tweet.id)
+
+
+async def check_update():
+    logs.info("开始检查更新")
+    users = UserDB.get_all()
+    nums = len(users)
+    for idx, user in enumerate(users):
+        try:
+            user_data = await get_user(user)
+            if user_data:
+                logs.info(f"获取 {user_data.name} 的数据成功，共 {len(user_data.tweets)} 条推文")
+                await send_check(user_data)
+            else:
+                logs.warning(f"获取 {user} 的数据失败，未知原因")
+        except UsernameNotFound:
+            logs.warning(f"获取 {user} 的数据失败，可能用户名已改变")
+            await send_username_changed(user)
+            UserDB.remove(user)
+        except Exception:
+            traceback.print_exc()
+        logs.info(f"处理完成，剩余 {nums - idx - 1} 个用户")
+    logs.info("检查更新完成")
diff --git a/init.py b/init.py
new file mode 100644
index 0000000..603b0d8
--- /dev/null
+++ b/init.py
@@ -0,0 +1,29 @@
+import pyrogram
+import httpx
+
+from defs.glover import api_id, api_hash
+from scheduler import scheduler
+from logging import getLogger, INFO, ERROR, StreamHandler, basicConfig
+from coloredlogs import ColoredFormatter
+
+# Enable logging
+logs = getLogger("T2G")
+logging_format = "%(levelname)s [%(asctime)s] [%(name)s] %(message)s"
+logging_handler = StreamHandler()
+logging_handler.setFormatter(ColoredFormatter(logging_format))
+root_logger = getLogger()
+root_logger.setLevel(ERROR)
+root_logger.addHandler(logging_handler)
+basicConfig(level=INFO)
+logs.setLevel(INFO)
+
+if not scheduler.running:
+    scheduler.start()
+
+bot = pyrogram.Client(
+    "bot", api_id=api_id, api_hash=api_hash, plugins=dict(root="plugins")
+)
+headers = {
+    "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/102.0.5005.72 Safari/537.36"
+}
+request = httpx.AsyncClient(timeout=60.0, headers=headers)
diff --git a/main.py b/main.py
new file mode 100644
index 0000000..4e96cfb
--- /dev/null
+++ b/main.py
@@ -0,0 +1,11 @@
+from pyrogram import idle
+
+from init import logs, bot
+
+
+if __name__ == "__main__":
+    logs.info("连接服务器中。。。")
+    bot.start()
+    logs.info("运行成功！")
+    idle()
+    bot.stop()
diff --git a/plugins/manage_user.py b/plugins/manage_user.py
new file mode 100644
index 0000000..e9dac75
--- /dev/null
+++ b/plugins/manage_user.py
@@ -0,0 +1,67 @@
+import re
+
+from pyrogram import filters
+from pyrogram.types import Message
+
+from init import bot
+
+from defs.sqlite import UserDB
+from defs.glover import owner
+
+
+@bot.on_message(filters=filters.command("add_user") & filters.user(owner))
+async def add_user(_, message: Message):
+    try:
+        username = message.text.split(" ")[1]
+    except IndexError:
+        await message.reply("请指定用户名！")
+        return
+    if UserDB.check(username):
+        await message.reply("该用户添加过了！")
+        return
+    UserDB.add(username)
+    await message.reply("添加成功！")
+
+
+@bot.on_message(filters=filters.regex(r"https://twitter.com/(.*)") & filters.chat(owner))
+async def add_user_regex(_, message: Message):
+    try:
+        username = re.findall(r"https://twitter.com/(.*)", message.text)[0]
+        if not username:
+            raise IndexError
+    except IndexError:
+        await message.reply("请指定用户名！")
+        return
+    if UserDB.check(username):
+        await message.reply("该用户添加过了！")
+        return
+    UserDB.add(username)
+    await message.reply(f"添加 {username} 成功！")
+
+
+@bot.on_message(filters=filters.command("del_user") & filters.user(owner))
+async def del_user(_, message: Message):
+    try:
+        username = message.text.split(" ")[1]
+    except IndexError:
+        await message.reply("请指定用户名！")
+        return
+    if not UserDB.check(username):
+        await message.reply("该用户未添加！")
+        return
+    UserDB.remove(username)
+    await message.reply(f"删除 {username} 成功！")
+
+
+@bot.on_message(filters=filters.command("list_user") & filters.user(owner))
+async def list_user(_, message: Message):
+    users = UserDB.get_all()
+    if not users:
+        await message.reply("列表为空！")
+        return
+
+    def format_user(username: str):
+        return f'<a href="https://twitter.com/{username}">{username}</a>'
+
+    text = "\n".join([format_user(user) for user in users])
+    await message.reply(f"用户列表：\n{text}")
diff --git a/plugins/ping.py b/plugins/ping.py
new file mode 100644
index 0000000..fc50303
--- /dev/null
+++ b/plugins/ping.py
@@ -0,0 +1,8 @@
+from init import bot
+
+from pyrogram import filters
+
+
+@bot.on_message(filters=filters.command("ping"))
+async def ping(_, message):
+    await message.reply("pong")
diff --git a/plugins/update.py b/plugins/update.py
new file mode 100644
index 0000000..f093360
--- /dev/null
+++ b/plugins/update.py
@@ -0,0 +1,21 @@
+from pyrogram.types import Message
+
+from defs.glover import owner
+from init import bot
+from scheduler import scheduler
+
+from pyrogram import filters
+
+from defs.update import check_update
+
+
+@bot.on_message(filters=filters.command("check_update") & filters.user(owner))
+async def update_all(_, message: Message):
+    msg = await message.reply("开始检查更新！")
+    await check_update()
+    await msg.edit("检查更新完毕！")
+
+
+# @scheduler.scheduled_job("cron", minute="*/30", id="update_all")
+# async def update_all_30_minutes():
+#     await check_update()
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..24bc539
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,10 @@
+pyrogram==2.0.106
+tgcrypto==1.2.5
+httpx==0.24.1
+feedparser
+beautifulsoup4
+lxml
+coloredlogs
+apscheduler
+sqlitedict
+pydantic
diff --git a/scheduler.py b/scheduler.py
new file mode 100644
index 0000000..148c970
--- /dev/null
+++ b/scheduler.py
@@ -0,0 +1,5 @@
+from apscheduler.schedulers.asyncio import AsyncIOScheduler
+
+scheduler = AsyncIOScheduler(timezone="Asia/ShangHai")
+if not scheduler.running:
+    scheduler.start()