diff --git a/tools/msg_tool.py b/tools/msg_tool.py
new file mode 100644
index 00000000..d7dbf56c
--- /dev/null
+++ b/tools/msg_tool.py
@@ -0,0 +1,358 @@
+import argparse
+import struct
+import os
+import re
+
+CHAR_MAP = [
+    "¡", "¿", "Ä", "À", "Á", "Â", "Ã", "Å", "Ç", "È", "É", "Ê", "Ë", "Ì", "Í", "Î",
+    "Ï", "Ð", "Ñ", "Ò", "Ó", "Ô", "Õ", "Ö", "Ø", "Ù", "Ú", "Û", "Ü", "ß", "Þ", "à",
+    " ", "!", "\"", "á", "â", "%", "&", "'", "(", ")", "~", "♥", ",", "-", ".", "♪",
+    "0", "1", "2", "3", "4", "5", "6", "7", "8", "9", ":", "🌢", "<", "=", ">", "?",
+    "@", "A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N", "O",
+    "P", "Q", "R", "S", "T", "U", "V", "W", "X", "Y", "Z", "ã", "💢", "ä", "å", "_",
+    "ç", "a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o",
+    "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z", "è", "é", "ê", "ë", "\u007f",
+    "�", "ì", "í", "î", "ï", "•", "ð", "ñ", "ò", "ó", "ô", "õ", "ö", "⁰", "ù", "ú",
+    "ー", "û", "ü", "ý", "ÿ", "þ", "Ý", "¦", "§", "ḏ", "ṉ", "‖", "µ", "³", "²", "¹", # note that a̱ and o̱ had to be changed because they're actually two characters in unicode. a̱ -> ḏ | o̱ -> ṉ
+    "¯", "¬", "Æ", "æ", "„", "»", "«", "☀", "☁", "☂", "🌬", "☃", "∋", "∈", "/", "∞",
+    "○", "🗙", "□", "△", "+", "⚡", "♂", "♀", "🍀", "★", "💀", "😮", "😄", "😣", "😠", "😃",
+    "×", "➗", "🔨", "🎀", "✉", "💰", "🐾", "🐶", "🐱", "🐰", "🐦", "🐮", "🐷", "\n", "🐟", "🐞",
+    ";", "#", "\u00d2", "\u00d3", "⚷", "\u00d5", "\u00d6", "\u00d7", "\u00d8", "\u00d9", "\u00da", "\u00db", "\u00dc", "Ỳ", "ꟓ", "\u00df",
+    "\u00e0", "\u00e1", "\u00e2", "\u00e3", "\u00e4", "\u00e5", "\u00e6", "\u00e7", "\u00e8", "\u00e9", "\u00ea", "\u00eb", "\u00ec", "\u00ed", "\u00ee", "\u00ef",
+    "\u00f0", "\u00f1", "\u00f2", "\u00f3", "\u00f4", "\u00f5", "\u00f6", "÷", "\u00f8", "\u00f9", "\u00fa", "\u00fb", "\u00fc", "\u00fd", "\u00fe", "\u00ff"
+]
+
+CONT_SIZES = [
+    2, 2, 2, 3, 2, 5, 2, 2, 5, 5, 5, 5, 5, 2, 4, 4,
+    4, 4, 4, 6, 8, 10, 6, 8, 10, 2, 2, 2, 2, 2, 2, 2,
+    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+    2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+    6, 3, 3, 3, 3, 2, 4, 4, 3, 3, 3, 2, 2, 2, 2, 2,
+    2, 2, 2, 6, 3, 3, 4, 3, 2, 2, 6, 2, 2, 3, 3, 3,
+    3, 2, 2, 2, 2, 2, 2, 4, 4, 12, 14
+]
+
+COMMANDS = [
+    'MSGEND',
+    'MSGCONTINUE',
+    'MSGCLEAR',
+    'PAUSE',
+    'BTN',
+    'TEXTCOLOR',
+    'ABLECANCEL',
+    'UNABLECANCEL',
+    'DEMOPLR',
+    'DEMONPC0',
+    'DEMONPC1',
+    'DEMONPC2',
+    'DEMONPCQST',
+    'OPENCHOICE',
+    'SETFORCEMSG',
+    'SETNEXTMSG0',
+    'SETNEXTMSG1',
+    'SETNEXTMSG2',
+    'SETNEXTMSG3',
+    'SETNEXTMSGRND2',
+    'SETNEXTMSGRND3',
+    'SETNEXTMSGRND4',
+    'SETSELSTR2',
+    'SETSELSTR3',
+    'SETSELSTR4',
+    'FORCENEXT',
+    'STR_PLAYERNAME',
+    'STR_TALKNAME',
+    'STR_TAIL',
+    'STR_YEAR',
+    'STR_MONTH',
+    'STR_WEEK',
+    'STR_DAY',
+    'STR_HOUR',
+    'STR_MIN',
+    'STR_SEC',
+    'STR_FREE0',
+    'STR_FREE1',
+    'STR_FREE2',
+    'STR_FREE3',
+    'STR_FREE4',
+    'STR_FREE5',
+    'STR_FREE6',
+    'STR_FREE7',
+    'STR_FREE8',
+    'STR_FREE9',
+    'STR_DETERMINATION',
+    'STR_COUNTRYNAME',
+    'STR_RNDNUM',
+    'STR_ITEM0',
+    'STR_ITEM1',
+    'STR_ITEM2',
+    'STR_ITEM3',
+    'STR_ITEM4',
+    'STR_FREE10',
+    'STR_FREE11',
+    'STR_FREE12',
+    'STR_FREE13',
+    'STR_FREE14',
+    'STR_FREE15',
+    'STR_FREE16',
+    'STR_FREE17',
+    'STR_FREE18',
+    'STR_FREE19',
+    'STR_MAIL',
+    'LUCK_NEUTRAL',
+    'LUCK_RELATIONSHIP',
+    'LUCK_UNPOPULAR',
+    'LUCK_BAD',
+    'LUCK_MONEY',
+    'LUCK_GOODS',
+    'LUCK_6',
+    'LUCK_7',
+    'LUCK_8',
+    'LUCK_9',
+    'MSGCONTENTS_NORMAL',
+    'MSGCONTENTS_ANGRY',
+    'MSGCONTENTS_SAD',
+    'MSGCONTENTS_FUN',
+    'MSGCONTENTS_SLEEPY',
+    'COLORCHARS',
+    'SNDCUT',
+    'LINEOFS',
+    'LINETYPE',
+    'CHARSCALE',
+    'BTN2',
+    'BGMMAKE',
+    'BGMDELETE',
+    'MSGTIMEEND',
+    'SNDTRGSYS',
+    'LINESCALE',
+    'SNDNOPAGE',
+    'VOICETRUE',
+    'VOICEFALSE',
+    'SELNOB',
+    'GIVEOPEN',
+    'GIVECLOSE',
+    'MSGCONTENTS_GLOOMY',
+    'SELNOBCLOSE',
+    'SETNEXTMSGRNDSECTION',
+    'AGBDUMMY0',
+    'AGBDUMMY1',
+    'AGBDUMMY2',
+    'SPACE',
+    'AGBDUMMY3',
+    'AGBDUMMY4',
+    'MALEFEMALECHK',
+    'AGBDUMMY5',
+    'AGBDUMMY6',
+    'AGBDUMMY7',
+    'AGBDUMMY8',
+    'AGBDUMMY9',
+    'AGBDUMMY10',
+    'STR_ISLANDNAME',
+    'SETCURSORJUST',
+    'CLRCUSRORJUST',
+    'CUTARTICLE',
+    'CAPTIALIZE',
+    'STR_AMPM',
+    'SETNEXTMSG4',
+    'SETNEXTMSG5',
+    'SETSELSTR5',
+    'SETSELSTR6'
+]
+
+def decode_control_code(ba: bytearray, idx: int):
+    if ba[idx] != 0x7F:
+        raise ValueError("First character must be 0x7F")
+    cont_type = ba[idx + 1]
+    if cont_type >= len(CONT_SIZES):
+        raise ValueError(f"Invalid control code id {cont_type:02X}")
+    cont_size = CONT_SIZES[cont_type]
+    if len(ba) < idx + cont_size:
+        raise ValueError(f"Bytearray is not large enough for control code {cont_type:02X}")
+
+    cmd = COMMANDS[cont_type]
+    if cmd is not None:
+        if cont_size > 2:
+            hex_values = ' '.join('{:02X}'.format(b) for b in ba[(idx + 2):(idx + CONT_SIZES[cont_type])])
+            return f"<<{cmd} [{hex_values}]>>", cont_size
+        else:
+            return f"<<{cmd}>>", cont_size
+    else:
+        hex_values = ' '.join('{:02X}'.format(b) for b in ba[(idx + 2):(idx + CONT_SIZES[cont_type])])
+        return f"<<Control Code [{cont_type:02X}] [{hex_values}]>>", cont_size
+
+def decode_entry(ba: bytearray, start: int, end: int, idx: int):
+    parts = [f'[[ENTRY {idx} START]]\n']  # Use a list to collect string parts
+    i = start
+    while i < end:
+        char = ba[i]
+        if char == 0x7F:
+            cont_str, cont_size = decode_control_code(ba, i)
+            parts.append(cont_str)
+            i += cont_size
+        else:
+            parts.append(CHAR_MAP[ba[i]])
+            i += 1
+    parts.append('\n\n')
+    return ''.join(parts)  # Join the parts into a final string at the end
+
+def decode_file(data_path: str, table_path: str, out_path: str):
+    idx = 0
+    last_end = 0
+    output_buffer = []
+
+    with open(data_path, 'rb') as df, open(table_path, 'rb') as tf, open(out_path, 'w') as of:
+        while True:
+            bytes = tf.read(4)
+            if not bytes:
+                break
+            end = struct.unpack('>I', bytes)[0]
+            if end != 0:
+                size = end - last_end
+                last_end = end
+                data = bytearray(df.read(size))
+                decoded_str = decode_entry(data, 0, size, idx)
+                output_buffer.append(decoded_str)
+
+            idx += 1
+            # Write buffer content to file to reduce write calls
+            if len(output_buffer) >= 8192:
+                of.write(''.join(output_buffer))
+                output_buffer.clear()
+
+        # Write remaining buffer content to file
+        if output_buffer:
+            of.write(''.join(output_buffer))
+
+# Function to convert a hex string to a list of integers
+def convert_hex_string_to_ints(hex_str):
+    # Remove spaces and convert to upper case
+    clean_str = hex_str.replace(" ", "").upper()
+    # Convert every two characters to an integer
+    return [int(clean_str[i:i+2], 16) for i in range(0, len(clean_str), 2)]
+
+# pre-compiled regex patterns
+cmd_pattern = re.compile(r"^([\w\s]+)")
+arg_pattern = re.compile(r"\[([0-9A-Fa-f\s]*)\]")
+def encode_control_code(cont_code_str: str, start_idx: int = 0, end_idx: int = None):
+    sliced_str = cont_code_str[start_idx:end_idx] # bad but necessary in python
+    cmd_match = cmd_pattern.match(sliced_str)
+    if not cmd_match:
+        raise ValueError("Missing command in control code!")
+
+    cmd = cmd_match.group(1).strip()
+    args = arg_pattern.findall(sliced_str)
+
+    cmd_idx = COMMANDS.index(cmd)
+    arg_list = [byte for hex_str in args for byte in convert_hex_string_to_ints(hex_str)]
+    return cmd_idx, arg_list
+
+def encode_entry(entry: str):
+    ba = bytearray()
+    i = 0
+    max = len(entry)
+    while i < max:
+        char = entry[i]
+        if char == '<' and i < max - 1 and entry[i + 1] == '<':
+            start = i + 2
+            end = start
+            found = False
+            while end < max:
+                if entry[end] == '>' and end < max - 1 and entry[end + 1] == '>':
+                    found = True
+                    break
+                end += 1
+            if found:
+                cmd_idx, arg_list = encode_control_code(entry, start, end)
+                cmd_size = CONT_SIZES[cmd_idx]
+                if len(arg_list) != cmd_size - 2:
+                    raise ValueError(f"Expected args of length {cmd_size - 2} for command {COMMANDS[cmd_idx]}, but got {len(arg_list)}")
+                ba.append(0x7F)
+                ba.append(cmd_idx)
+                ba.extend(arg_list)
+                i = end + 2
+                continue
+        
+        ba.append(CHAR_MAP.index(char))
+        i += 1
+    return ba
+
+def encode_file(file_path: str, data_path: str, table_path: str, data_size: int=-1, table_size: int=-1):
+    entries = {}
+    current_entry = None
+    recording = False
+
+    with open(file_path, 'r') as tf, open(data_path, 'wb') as df, open(table_path, 'wb') as tabf:
+        idx = -1
+        for line in tf:
+            l = line.strip()
+
+            # Check for entry start
+            if l.startswith("[[ENTRY") and l.endswith("START]]"):
+                entry_index = l.split()[1]  # Assuming the format [[ENTRY X START]]
+                current_entry = entry_index
+                entries[current_entry] = []
+                recording = True
+                continue
+
+            # Check for entry end
+            if line.find("<<MSGEND>>") != -1 or line.find("<<MSGTIMEEND") or line.find("<<MSGCONTINUE>>"):
+                entries[current_entry].append(line)
+                recording = False
+                continue
+
+            # Record lines if within an entry and not empty
+            if recording and line:
+                entries[current_entry].append(line)
+
+        #end_ofs = 0
+        for entry in entries:
+            entries[entry] = encode_entry(''.join(entries[entry]).rstrip())
+            df.write(entries[entry])
+            tabf.write(struct.pack('>I', df.tell()))
+        
+        if data_size > 0:
+            data_remain = data_size - df.tell()
+            if data_remain > 0:
+                df.write(b'\x00' * data_remain)
+
+        if table_size > 0:
+            table_remain = table_size - tabf.tell()
+            if table_remain > 0:
+                tabf.write(b'\x00' * table_remain)
+        
+    return entries
+
+def main():
+    parser = argparse.ArgumentParser(description='Pack or dump Animal Crossing text files.')
+    parser.add_argument('-m', help="The mode to run. Valid arguments are un[pack].")
+    parser.add_argument('path', help='The path of the source file.')
+    parser.add_argument('out', help='The path of the destination file.')
+    parser.add_argument('--data_size', help='Optional hexadecimal padded size for the data file.', required=False)
+    parser.add_argument('--table_size', help='Optional hexadecimal padded size for the table file.', required=False)
+
+    args = parser.parse_args()
+    if args.m.lower() == "pack":
+        # Create *_table.bin path
+        dir_name, file_name = os.path.split(args.out)
+        name, ext = os.path.splitext(file_name)
+        new_file_name = f"{name}_table{ext}"
+
+        # encode
+        encode_file(args.path, args.out, os.path.join(dir_name, new_file_name), int(args.data_size, 16) if args.data_size != None else -1, int(args.table_size, 16) if args.table_size != None else -1)
+    elif args.m.lower() == "unpack":
+        # Search for *_table.bin
+        dir_name, file_name = os.path.split(args.path)
+        name, ext = os.path.splitext(file_name)
+        new_file_name = f"{name}_table{ext}"
+        table_path = os.path.join(dir_name, new_file_name)
+
+        if not os.path.exists(table_path):
+            raise Exception(f'Couldn\'t find a valid table path. Please ensure {new_file_name} exists!')
+        
+        # decode
+        decode_file(args.path, table_path, args.out)
+    else:
+        raise Exception(f'Invalid mode! Please use -m un[pack]')
+
+if __name__ == '__main__':
+    main()