decode-packet

---

name: decode-packet description: Decode MapleStory2 packet hex bytes interactively. Supports typed field parsing and little-endian conversions. argument-hint: [mode] [args...] allowed-tools: Bash(node -e *), AskUserQuestion

MapleStory2 Packet Decoder

Arguments received: $ARGUMENTS

---

Quick reference

Mode	Shorthand example
Parse hex as typed fields	/decode-packet hex "A6 98 14 00 00 01 04" int byte byte byte
Decimal → little-endian hex	/decode-packet to-hex 23000054
Little-endian hex → decimal	/decode-packet from-hex "F6 F3 5E 01"

---

Step 1 — Detect mode

Parse $ARGUMENTS:

• If it starts with hex → Mode A: typed field parse
• If it starts with to-hex → Mode B: decimal to little-endian hex
• If it starts with from-hex → Mode C: little-endian hex to decimal
• If empty or unrecognized → ask the user which mode they want using AskUserQuestion

---

Mode A — Hex bytes → typed fields

Usage: /decode-packet hex "<hex string>" <type> [type ...]

Supported types: byte, bool, short, int, uint, long, ulong, float, string

• Sizes: byte/bool=1, short=2, int/uint/float=4, long/ulong=8, string=2+N*2 (UTF-16LE, length-prefixed short)
• All multi-byte types are little-endian

Extract the hex string (first quoted arg after hex) and the list of types, then run:

node -e "
const hex = '<HEX_STRING>';
const types = [<TYPES_ARRAY>];
const buf = Buffer.from(hex.replace(/\\s+/g, ''), 'hex');
let offset = 0;
for (const type of types) {
  let value, size;
  if (type === 'byte') { value = buf.readUInt8(offset); size = 1; }
  else if (type === 'bool') { value = buf.readUInt8(offset) !== 0 ? 'true' : 'false'; size = 1; }
  else if (type === 'short') { value = buf.readInt16LE(offset); size = 2; }
  else if (type === 'int') { value = buf.readInt32LE(offset); size = 4; }
  else if (type === 'uint') { value = buf.readUInt32LE(offset); size = 4; }
  else if (type === 'long') { value = buf.readBigInt64LE(offset).toString(); size = 8; }
  else if (type === 'ulong') { value = buf.readBigUInt64LE(offset).toString(); size = 8; }
  else if (type === 'float') { value = buf.readFloatLE(offset).toFixed(6); size = 4; }
  else if (type === 'string') {
    const len = buf.readUInt16LE(offset);
    value = '\"' + buf.slice(offset + 2, offset + 2 + len * 2).toString('utf16le') + '\"';
    size = 2 + len * 2;
  } else { console.error('Unknown type: ' + type); process.exit(1); }
  const hexSlice = buf.slice(offset, offset + size).toString('hex').toUpperCase().match(/../g).join(' ');
  console.log('[offset=' + offset + '] ' + type + ': ' + value + '  (' + hexSlice + ')');
  offset += size;
}
if (offset < buf.length) {
  const remaining = buf.slice(offset).toString('hex').toUpperCase().match(/../g).join(' ');
  console.log('\\n[' + (buf.length - offset) + ' bytes remaining]: ' + remaining);
}
"

Example: /decode-packet hex "A6 98 14 00 00 01 04" int byte byte → parses 4+1+1 bytes: ObjectId, Command, Count

---

Mode B — Decimal → little-endian hex

Usage: /decode-packet to-hex <number> [width]

Width defaults to 4 bytes (Int). Use 8 for Long.

node -e "
const n = BigInt('<NUMBER>');
const width = <WIDTH>;  // 1, 2, 4, or 8
const buf = Buffer.alloc(width);
if (width === 1) buf.writeUInt8(Number(n));
else if (width === 2) buf.writeUInt16LE(Number(n));
else if (width === 4) buf.writeUInt32LE(Number(n));
else if (width === 8) buf.writeBigUInt64LE(n);
const spaced = buf.toString('hex').toUpperCase().match(/../g).join(' ');
console.log('Decimal : ' + n);
console.log('Hex (BE): 0x' + n.toString(16).toUpperCase().padStart(width * 2, '0'));
console.log('LE bytes: ' + spaced + '  ← use this with --search-hex');
"

---

Mode C — Little-endian hex → decimal

Usage: /decode-packet from-hex "<hex bytes>" (auto-detects width from byte count)

node -e "
const hex = '<HEX_STRING>';
const buf = Buffer.from(hex.replace(/\\s+/g, ''), 'hex');
if (buf.length === 1) console.log('byte  : ' + buf.readUInt8(0));
else if (buf.length === 2) {
  console.log('short : ' + buf.readInt16LE(0));
  console.log('ushort: ' + buf.readUInt16LE(0));
} else if (buf.length === 4) {
  console.log('int   : ' + buf.readInt32LE(0));
  console.log('uint  : ' + buf.readUInt32LE(0));
  console.log('float : ' + buf.readFloatLE(0).toFixed(6));
} else if (buf.length === 8) {
  console.log('long  : ' + buf.readBigInt64LE(0).toString());
  console.log('ulong : ' + buf.readBigUInt64LE(0).toString());
} else {
  console.log('(raw ' + buf.length + ' bytes — showing all interpretations)');
  console.log('uint32 from offset 0: ' + buf.readUInt32LE(0));
}
"

---

No arguments — ask the user

If $ARGUMENTS is empty or not one of the recognized modes, use AskUserQuestion to ask:

"What would you like to decode?" with options:

1. Parse hex bytes as typed fields (e.g. int, byte, short…)
2. Convert a decimal number → little-endian hex (for --search-hex)
3. Convert little-endian hex → decimal

Then ask a follow-up for the hex/number input once you know which mode they want.

---

Notes

• All values are little-endian (same as MapleStory2 protocol)

---

Data type cheat sheet

C# type	Bytes	Node.js read method
byte	1	readUInt8
bool	1	readUInt8 !== 0
short	2	readInt16LE
int	4	readInt32LE
uint	4	readUInt32LE
long	8	readBigInt64LE
float	4	readFloatLE
string	2+N*2	len=readUInt16LE, then toString('utf16le')