summary: the fastest path from signed up to live algo — what you need, what to do in order, and which article to open for each step.

this article covers: the 5-step path to a live algo, the broker contract-symbol rule that traps most new users, the position-sizing input that controls how many contracts each trade fires, and where to go next when something breaks.

if you already have an algo running and something's off, skip to → my algo didn't trade — start here.

prefer to watch first? Andre walks through the entire flow end-to-end — optimizer, dashboard, broker connection, strategy creation, and the TradingView alert — in the video below. the full 11-minute walkthrough also lives on start here in the algos sidebar (edgeful.com/algos-automation/start-here): pick video onboarding, or checklist onboarding if you'd rather click through step by step. the written steps below cover everything in the video plus the data and contract-symbol requirements it doesn't go deep on.

want to start a step earlier, at picking the algo? this second video walks through finding a top-performing algo, optimizing its settings, and getting it live: the optimizer, the prop-firm pass-rate sim, and copying your settings into TradingView with the Chrome extension.

5 things need to be in place before you touch the dashboard:

if any of those 5 aren't ready, fix them first — the rest of the setup won't hold together without them.

the #1 cause of silent algo failures is the wrong contract symbol on your TradingView alert. each broker platform expects a different symbol — get it wrong and orders either get rejected or never route. the rule:

rule of thumb: Tradovate / NinjaTrader = front-month. ProjectX = continuous.

that rule is for live trading. for backtesting, use the continuous contract (e.g., NQ1!, ES1!) whatever your broker.

on Tradovate / NinjaTrader, futures contracts roll quarterly — you'll need to update the chart symbol and recreate the alert each quarter. ProjectX users on continuous symbols don't have this problem.

your algo fires a signal on TradingView → TradingView fires an alert through the strategy webhook to edgeful → edgeful routes the order to your broker → your broker executes the trade.

each link is independent. when something goes wrong, you narrow down which link broke — that's what the troubleshooting flow below is built for. for the under-the-hood reference (how TP and SL exits are delivered, where slippage comes from), see → how the algo automation chain works.

step 1 — get the algo indicators. edgeful's algo indicators (ORB, IB, engulfing) are invite-only TradingView scripts. in edgeful, open algos in the left sidebar and click algo access. enter your TradingView username (not your display name or email: it's the one in your profile URL, tradingview.com/u/[your-username]) and click update. then in TradingView, open Indicators and select Invite-only in the left sidebar of the indicators panel. the edgeful algo indicators should be there.

not showing up? close every TradingView tab and window (don't just refresh) and reopen it, and double-check you entered the right username. still missing: → TradingView indicators: access, updates, and troubleshooting or my algo indicators aren't showing up in TradingView.

step 2 — connect your broker. in edgeful, open the algo dashboard from algos → dashboard in the left sidebar. click connect Tradovate/NinjaTrader or connect ProjectX, log into your broker, approve the connection, and confirm the status shows connected in green.

full walkthrough — including 2FA pitfalls and the ProjectX errorCode=3 fix: → using the algo dashboard

step 3 — create a strategy. in the same dashboard, click create strategy and fill in all 4 fields: name (e.g. "NQ ORB SIM"), description, broker account, and entry order type (market or limit). the create button stays disabled until every field is filled. point this first strategy at a SIM/demo account, since you're testing the chain, not your bankroll.

the strategy appears in the table with status enabled (green), plus its own webhook URL and JSON message. copy both, you need them in step 4. use the copy icon (the dual-box icon) next to "view webhook" to grab the full URL. never copy from the masked preview, it's incomplete.

one strategy per broker account. 3 accounts = 3 strategies = 3 separate TradingView alerts. for the multi-account, multi-prop-firm scenarios matrix, see → using edgeful algos with prop firms and multiple accounts.

step 4 — build your TradingView alert. add your edgeful algo indicator to a chart (use regular candlesticks, not heikin ashi, which cause order rejections), set contracts per trade in the properties tab of the indicator's settings — it's the number on the default order size line, and the type next to it must be quantity, not % of equity, or the algo never places a trade (this is the only place position size is set — full breakdown: → position sizing and how edgeful uses strategy.position_size), open the alert dialog, paste the webhook URL into the Webhook URL field, toggle the webhook notification on (the URL alone doesn't fire anything), and paste the JSON message into the message field. match your chart timeframe to your algo's timeframe, set the alert to trigger on your strategy's conditions (not a price level or indicator), and set the expiration as far out as your plan allows.

full walkthrough, including what to do after a settings or script update: → setting up TradingView alerts for algo automation

step 5 — verify it's working. run the 5-minute SIM pre-flight before trusting the algo with real money. with your strategy on a SIM/demo account at a supported broker (a Tradovate demo, or an eval/sim account on a ProjectX-based prop firm), wait for a real signal to fire. TradingView's paper trading mode won't fire automation, so it can't test the chain. then confirm your trade shows up in the trade log (algos → trade log, edgeful.com/algos-automation/trade-log) and went through at your broker. the trade log shows one day at a time, so use the strategy filter and the status chips (all, working, closed, rejected, cancelled) to find your trade. when the trade exits (TP or SL), the exit shows up in the trade log too, and the position closes at the broker.

full walkthrough: → test your webhook before going live. for what each log status means, see → understanding the algo trade log.

don't jump straight to your target size on a live account. the right order:

when switching from SIM to live, create a new strategy pointed at your live account, copy its unique webhook URL, and create a new TradingView alert with that URL. don't reuse the SIM webhook — each strategy has its own URL, and the SIM webhook keeps routing to SIM. before the first live signal, make sure your chart is on the right contract for your broker (front-month for Tradovate / NinjaTrader, continuous for ProjectX) and that your broker shows the correct contract.

once you're live, any change to your edgeful algo indicator settings (session, ORB size, TPs, SLs, contracts per trade — anything) means you need to delete and recreate your TradingView alert. the alert is locked to the indicator settings at the time it was created. your webhook URL stays the same — only the alert needs rebuilding.

full workflow when changes come from the algo optimizer or a manual tweak: → applying algo settings changes — what to do after you optimize and make algo setting changes.

a few things to stay on top of once you're live:

already code your own strategies in TradingView? you can automate them through edgeful the same way you automate an edgeful algo: connect your broker, create a strategy in the dashboard, then create the TradingView alert on your strategy with that strategy's webhook URL.

the key difference: you must include the alert message. paste the JSON message from the dashboard into your alert's message field, exactly as shown. that message is how edgeful reads the trade, so without it the alert fires but the order won't route.

want to see it done? the pro tip on the algo dashboard (algos → dashboard) links to a walkthrough video.

one limit to know: the algo optimizer and algo templates only work with edgeful's algos. you can't upload your own Pine script to the optimizer.

when a trade doesn't execute, the problem is always in one of 3 places — the algo didn't find a setup, TradingView didn't fire the alert, or the alert didn't reach edgeful through the webhook.

start here: → my algo didn't trade — start here. the article walks you through 3 diagnostic questions and points you to the right fix.

if you're already looking at an exact error string in the trade log — "rejected", "failed", "contract not found", "cannot sell when position is 0", etc. — jump straight to the → algo error message decoder. every common error is indexed there with a direct fix.

my algo fired late / at the wrong price — what's wrong?

almost always delayed chart data. TradingView shows CME futures on a ~10-minute delay by default, so alerts fire late and entries look "wrong" even though the algo is working correctly. the fix is a CME real-time data package on TradingView (on top of a paid plan) — see enabling real-time CME futures data on TradingView. also confirm your chart timeframe matches your algo timeframe, since alerts only fire on bar close.

what's the difference between standard and holdout validation in the optimizer?

standard runs on the full backtest period — fast, uses all data. holdout splits the period 80/20 and re-runs the winners on the held-back 20% to test out-of-sample. use holdout to verify winners before trusting them. see algo optimizer.

why isn't the #1 ranked combination the one with the highest P&L?

the composite score weighs P&L together with drawdown, consistency, win rate, profit factor, and trade count — not just P&L. scan the whole top 20 rather than only clicking row 1.

what does the "low overfit risk" badge on the strategy health card actually check?

7 individual pass/fail checks: profit factor, win rate, concentration risk, return consistency, performance decay, drawdown/profit ratio, and stability check. all 7 passing earns the low overfit risk badge.

can I upload my own Pine script to the optimizer?

no, you can't upload your own Pine script to the optimizer. it works with edgeful's built-in algos (engulfing 2 TP included). you can still automate your own strategy through edgeful, though. see automating your own TradingView strategy above.

why isn't my algo trading?

start with my algo didn't trade — start here. most common cause: TradingView alert isn't set, the webhook isn't configured, or your indicator settings don't match the optimized output.

does edgeful support NinjaTrader, or is it TradingView only?

yes, as an execution broker. but automation is always driven from TradingView: the algo runs as a strategy on a TradingView chart, the chart fires an alert/webhook to edgeful, and edgeful routes the order to your broker. NinjaTrader is one of the supported execution brokers (alongside Tradovate and ProjectX), not an alternative automation platform. so you still need TradingView to run the algo; NinjaTrader is where the trade lands. see do edgeful algos require TradingView? (NinjaTrader vs. TradingView).

does TradingView need to stay open?

no. TradingView alerts run on their servers, so you don't need the browser or app open. you just need an active paid TradingView plan with your alerts set up.

can I use the algos manually instead of automating?

yes. you can use the indicators visually and manage entries yourself, so automation isn't required. the algo shows entry, TP, and SL levels on your chart whether or not you've connected a broker.

→ more algo questions in the central FAQ

spokes — the setup steps:

data requirement:

after you've optimized or changed settings:

read the trade log:

troubleshooting:

reference: