Indicators

Understanding Indicators

How your agent reads the market

Indicators are technical tools that transform raw price data into signals your agent can act on. They're the same tools human traders have used for decades — RSI, moving averages, MACD, Bollinger Bands — except your agent can run them across multiple timeframes simultaneously and interpret the results in context.

What Is an Indicator?

At its core, an indicator is a formula that takes price data in and produces a signal out. A moving average smooths out noise to show you the trend. RSI tells you if something is overbought or oversold. MACD shows you when momentum is shifting.

Human traders look at these on charts and make gut calls. Your agent does the same thing — except it does it mathematically, consistently, and without emotion.

Inputs — What the Agent Controls

Every indicator has inputs — parameters the AI can adjust when it calls the indicator. This is important: your agent doesn't just get a static number. It chooses how to run the indicator based on your strategy and the current situation.

Common inputs:

  • Period — How many candles to look back. A 9-period EMA is fast and reactive. A 200-period EMA is slow and shows the big picture. Your agent can request different periods in the same analysis to see both.
  • Timeframe — What candle size to analyze. Your agent can call the same indicator on the 5-minute chart for precision timing, then again on the 1-hour chart for trend confirmation. All in the same analysis cycle.
  • Thresholds — Levels like RSI's overbought (70) and oversold (30) lines, or Bollinger Band width. These can be tuned to match the current market.

The AI decides which inputs to use based on your strategy. If your strategy says "use the 5-minute chart for entries and the 1-hour chart for trend direction," the agent will call your indicators on both timeframes and combine the results.

Outputs — What the Agent Reads

When an indicator runs, it returns outputs — the actual data the AI uses to make decisions. These aren't just raw numbers. They're structured signals that give the agent a clear picture.

Example — what the agent sees when it calls EMA (on any asset — crypto, forex, stocks, or commodities):

  • ema_9: 47,250.23 — The current value of the 9-period EMA
  • ema_21: 47,180.45 — The current value of the 21-period EMA
  • ema_50: 46,920.10 — The current value of the 50-period EMA
  • ema_200: 45,100.88 — The current value of the 200-period EMA

From these four numbers, the agent can see: all EMAs are stacked bullish (9 > 21 > 50 > 200), price is in an uptrend, and the 9 EMA is pulling away from the 21 — momentum is building. This works the same whether you're trading BTC/USD, EUR/USD, AAPL, or Gold.

Example — what the agent sees when it calls RSI:

  • rsi: 28 — The current RSI value
  • oversold: true — RSI is below the oversold threshold
  • momentum: "moderate_down" — Selling pressure is easing

The agent reads all of this together, across every indicator you've attached, and makes a decision.

Multi-Timeframe Analysis

This is one of the most powerful features. Every indicator your agent has can be called on any timeframe — 1-minute, 5-minute, 15-minute, 1-hour, 4-hour, or daily.

Your agent doesn't just look at one chart. It can:

  • Check the daily chart to see the big-picture trend
  • Check the 1-hour chart to confirm momentum
  • Check the 5-minute chart to time a precise entry

All in a single analysis cycle. The strategy you write guides which timeframes the agent prioritizes. Write "use the 5-minute for entries and the 1-hour for trend" and the agent follows that.

This is how professional traders think — top-down, from the macro view to the micro view. Your agent does it automatically.

Multi-timeframe confluence = higher confidence. When the same signal shows up across multiple timeframes, it's a much stronger signal. Your agent understands this and weighs multi-timeframe agreement into its confidence score.

Next Steps