How does Agenticks AlgoAgent work?

Open /algoagent and describe the strategy, for example an NQ opening range breakout on 5-minute bars with the stop under the range, or attach a PDF of your rules. The agent writes exact entry, exit, sizing, session and risk rules and asks about anything ambiguous instead of guessing. It writes the strategy as Python and runs it over the included history. You get every trade, the equity curve, drawdown, win rate, profit factor, a regime breakdown and Monte Carlo resampling of the real trades. Ask whether it would pass a Topstep, FTMO or Apex evaluation, change a rule, or save it. Run it live: signals to Telegram or a paper account on Assistant, real orders at your own broker on Agent.

What can Agenticks AlgoAgent do?

Strategy building from plain English, no code Backtests on included futures, stock, crypto and forex data ICT and smart money concepts, opening range breakouts, VWAP and session setups turned into testable rules Monte Carlo resampling of the real trades Prop firm evaluation check on Topstep, FTMO or Apex rules, or your own target and drawdown Market studies: questions about how a market behaved, answered over real data TradingView Pine Script v6 indicators written from a description Live trading by signal, paper account or broker order

AlgoAgent Docs

What this does

AlgoAgent is the agent at the centre of Agenticks. You describe a trading strategy the way you would explain it to another trader, and it turns it into exact rules, writes them as Python code you can read, backtests them on the included market data, and hands back every trade. It also checks the result against prop firm evaluation rules, writes TradingView indicators, and takes a finished strategy live.

How it works

Open /algoagent and describe the strategy, for example an NQ opening range breakout on 5-minute bars with the stop under the range, or attach a PDF of your rules.
The agent writes exact entry, exit, sizing, session and risk rules and asks about anything ambiguous instead of guessing.
It writes the strategy as Python and runs it over the included history.
You get every trade, the equity curve, drawdown, win rate, profit factor, a regime breakdown and Monte Carlo resampling of the real trades.
Ask whether it would pass a Topstep, FTMO or Apex evaluation, change a rule, or save it.
Run it live: signals to Telegram or a paper account on Assistant, real orders at your own broker on Agent.

Capabilities

Strategy building from plain English, no code
Backtests on included futures, stock, crypto and forex data
ICT and smart money concepts, opening range breakouts, VWAP and session setups turned into testable rules
Monte Carlo resampling of the real trades
Prop firm evaluation check on Topstep, FTMO or Apex rules, or your own target and drawdown
Market studies: questions about how a market behaved, answered over real data
TradingView Pine Script v6 indicators written from a description
Live trading by signal, paper account or broker order

Who uses AlgoAgent?

Mostly futures day traders working NQ, MNQ, ES and gold on 1-minute to 15-minute bars, many of them on prop firm evaluations, along with stock and crypto traders and TradingView users who want indicators written for them.

Do I need my own market data?

You do not bring your own market history. Agenticks includes futures NQ, MNQ, ES, MES, GC, MGC, SI and SIL from 2016, 1,776 US stocks and ETFs from as early as 2000, 12 crypto pairs from 2017 and EURUSD from 2021, at 1m, 5m, 15m, 30m, 1h and 1d. See Included market data.

Which plan includes AlgoAgent?

Every new account starts with a 7-day trial. Assistant is $49.99 a month and includes the agent, backtests over up to 10 years of history, the prop firm check, indicators, Telegram signals and a paper account. Agent is $99.99 a month and adds real orders at your own broker, the full history, and Claude connected to your account. See Pricing.

Limits

AlgoAgent does not hand anybody a profitable strategy and makes no claim that it can. It shows every trade so you can judge the idea yourself, and most ideas do not survive that. A win rate alone says little: in one Agenticks test a strategy won 60.2% of its trades and still lost money, because the average loss was 1.5 times the average win. Backtests do not predict future results.