📘Syntax & Commands

This sectionsection documents the supported syntax, commands, and primitives available in the memejob strategy engine. It serves as a reference and is not intended to be read linearly.

Strategies generated via the LLM-assisted builder must still conform to the syntax and execution model described below.

Below are all the supported syntax, variables and functions available when building trading strategies using the WAKOS / memejob LUA Strategy Builder.

The environment is a sandboxed LUA DSL designed for safe, fast, deterministic backtesting and execution. It is not full LUA.

Not all commands are required for every strategy; most strategies use a small subset of the primitives listed below.

1. Execution Model (Important)

Understanding the execution model is critical before writing strategies.

• The strategy script is executed once per bar

• Scripts are stateless by default (no memory across bars)

• The engine manages: Positions and Order execution

• Your script expresses intent via return values

Return Values

return 1   -- Buy / Long / Add (DCA)
return -1  -- Exit or Short (engine-defined)
return 0   -- Do nothing / Hold

2. Supported Lua Syntax

The strategy builder supports a subset of LUA.

✅ Supported commands

• local variable declarations

• Numeric arithmetic: + - * /

• Comparisons: > < >= <= ==

• Boolean logic: and, or

• Conditional blocks: if / elseif / end

• Single-return strategy flow

Example:

local e21 = ema(21)

if close < e21 then
    return 1
elseif close > e21 then
    return -1
end

return 0

❌ Not Supported (by design)

• Loops (for, while)

• Functions

• Tables / arrays

• Persistent state across bars

• LUA standard libraries (math.*, string.*, etc.)

3. Core Market Data Variables

All market data variables are instances of ValueWithHistory.

They support:

• Direct access (close)

• Historical indexing (close[1], close[2])

Price Variables

Variable	Description
open	Opening price of current bar
high	Highest price of current bar
low	Lowest price of current bar
close	Closing price of current bar
volume	Trading volume of current bar

Example:

local current = close
local previous = close[1]

if close > close[1] then
    return 1
end

4. Derived Price Variables

Calculated prices derived from OHLC values.

Variable	Formula	Description
hl2	(high + low) / 2	Mid price
hlc3	(high + low + close) / 3	Typical price
hlcc4	(high + low + close + close) / 4	Close-weighted price

All support historical indexing.

Example:

if hlc3 > hlc3[1] then
    return 1
end

5. Bar Event Context (ev)

The global ev table provides context about the current bar event.

Fields

Field	Type	Description
ev.kind	string	Opened, Updated, or Closed
ev.bar_span	number	Bar interval in minutes

Example:

if ev.kind == "Closed" then
    return 0
end

6. Moving Average Indicators (Module MA)

EMA – Exponential Moving Average

ema(periods[, window="1m"])

Parameters:

• periods (number, required)

• window (string | number, optional)

Examples:

local e3 = ema(3)
local e21 = ema(21, "15m")
local e55 = ema(55, 60)

Asset-specific EMA:

local btc_ema = ema("BTC", 14, "15m")

---

SMA – Simple Moving Average

sma(window[, resolution="1m"])

Examples:

local sma_15m = sma("15m")
local sma_1h = sma("1h", "5m")

---

WMA – Weighted Moving Average

wma(periods, window, weights)

Examples:

local wma_val = wma(14, "15m", 14)

---

MACD – Moving Average Convergence Divergence

macd([small_periods=12[, large_periods=26[, window="1m"]]])

Examples:

local m = macd()
local m_fast = macd(9, 21, "15m")

Note: MACD always uses the primary asset.