SabrTrader Documentation

Alerts: Conditional Builder and AI-Generated Alerts

SabrTrader alerts are background monitors that watch a symbol and notify you when a defined condition is met. They range from a single price-level break to multi-condition logic that combines indicator plots, order-flow events such as delta, volume comparisons and session data, joined with AND/OR/NOT operators and a bars-based lookback window. The same alert system can also be driven by the built-in AI chat assistant, which can apply indicators to a chart, split overnight and regular-hours volume profiles, mark supply and demand zones, and then generate price-level alerts automatically from those indicator levels.

Overview: What SabrTrader Alerts Can Do

An alert in SabrTrader is a rule that the platform evaluates continuously against incoming market data for one symbol on one bar interval. When the rule becomes true, the platform delivers a notification through the channels you selected: email, Telegram, sound, popup, or the activity log.

There are two ways to create alerts, and this page covers both:

1. The manual alert builder. You open New > Alerts, click New Alert, and define conditions yourself. This scales from the simplest possible rule ("close greater than 500") to compound logic involving indicators, order-flow events, session comparisons and volume filters combined with Boolean operators.
2. The AI chat assistant. You describe what you want in plain language. The AI can apply indicators to the chart (for example the Volume Profile indicator), split the overnight session from the regular-hours session, draw supply and demand zones, and then create a set of price-level alerts derived from the levels those indicators produced. Nothing reaches the chart or the alert list until you press Apply.

Key properties of the alert engine

• Alerts do not require an open chart. The symbol you set an alert on does not need to be loaded in the current workspace. Alerts are stored and evaluated by the platform independently of your layout, which makes them the correct tool for watchlist monitoring of markets you are not actively watching on screen.
• Alerts are per-symbol and per-interval. Each alert names one instrument and one bar aggregation. Conditions are evaluated on bars of that aggregation.
• Alerts persist across workspaces. Closing a chart, switching workspaces or rearranging panels does not delete or disable an alert.
• Alert logic approaches strategy logic but is not a strategy. An alert tells you to look; it does not place, size or manage an order. When your rule set becomes a genuine trade plan with entries, exits and risk, move it to AlgoStudio Pro or the Auto Strategy Builder.

Opening and Using the Alert Manager

The alert manager is the window that lists every alert you own. Open it from the Control Center with New > Alerts.

What the list shows

• Name — the label you gave the alert. This text is what appears in the notification, so descriptive names such as GOOGL 15m level break 500 are far more useful than Alert 1.
• Symbol — the instrument the alert monitors.
• Interval / timeframe — the bar aggregation the conditions are evaluated on (time, tick, volume or range).
• Condition summary — a compressed description of the rule, for example Close crosses above 77.61.
• Enabled state — whether the alert is currently armed. Disabled alerts stay in the list but never fire.
• Notification channels — which delivery methods are attached.

Grouping: user alerts versus AI alerts

The manager separates alerts you created by hand from alerts generated by the AI assistant. This distinction matters operationally: AI level alerts are snapshots of indicator levels at the moment they were generated, so they are the group you clear out and regenerate when sessions roll. Your hand-built confluence alerts usually stay in place for weeks.

Actions available on a selected alert

• Edit — reopen the alert dialog to change the name, symbol, interval, conditions or channels. AI-generated alerts open in exactly the same dialog as manual ones.
• Enable / Disable — temporarily silence an alert without losing its configuration. Useful for alerts that only matter during certain sessions.
• Duplicate — copy an alert as the starting point for a variant, for example the same logic on a different symbol.
• Delete — remove the alert permanently.

Anatomy of the New Alert Dialog

Clicking New Alert opens a dialog with a header section (identity) and a body section (conditions and notifications).

Name

A free-text label. The name is carried into the email, Telegram message and popup, so write it to answer the question "what should I look at?" Examples: level break, AAPL 5m RSI+delta confluence, MES overnight VAH cross.

Symbol

Any tradable instrument your connections can resolve: equities such as GOOGL or AAPL, futures such as MES, forex pairs, crypto pairs. The symbol does not have to be open anywhere in your workspace. If a condition requires trade-by-trade data (any order-flow event), the connection supplying that symbol must deliver tick data rather than pre-aggregated bars.

Timeframe / interval

The bar aggregation on which conditions are evaluated. SabrTrader supports time-based bars (30-second, 5-minute, 15-minute), tick bars (for example 2000-tick), volume bars and range bars.

The interval is not cosmetic — it changes the meaning of every threshold in the alert:

Interval	What a "delta > 800" condition implies	Typical use
30-second	An intense, short burst of net aggressive buying — a genuine order-flow event	Order-flow triggers, absorption and initiative detection
5-minute	A moderate net imbalance accumulated over five minutes	Balanced confluence rules
15-minute	Often reached in routine conditions; a weak filter	Momentum and structure filters, not order-flow triggers
2000-tick	Depends entirely on instrument activity; the bar length varies with pace	Activity-normalised order-flow work

Rule of thumb: put order-flow conditions on fast intervals and momentum or structure conditions on slower intervals. If you need both, use two conditions with a bars lookback rather than forcing one interval to serve both purposes.

Bar close versus intrabar evaluation

A condition written against close is a bar-close condition: it can only be judged when the bar completes, so the alert fires at the end of the bar. A condition written against last price, high, low or a crosses above/below comparison can become true while the bar is still forming, giving an earlier but less confirmed trigger. Choose deliberately: bar-close conditions produce fewer, later, more reliable alerts; intrabar conditions produce earlier alerts that can un-happen before the bar closes.

Condition Source Types in Detail

Every condition row starts with a source dropdown. The source determines what data the left side of the comparison reads from. The row then exposes an operator (greater than, less than, equal to, crosses above, crosses below) and a right side (a constant, another series, or another plot).

(a) Price level

Compares a price series against a constant or another series.

• Series options: close, open, high, low, last price, typical price.
• Operators: greater than, less than, crosses above, crosses below.
• Example: Close > 500 on GOOGL 15-minute — fires on the first 15-minute close above 500.
• "Crosses" versus "greater than": > is true on every bar while price stays above the level and can re-fire repeatedly. Crosses above is true only on the transition bar, which is usually what you want for a level break.
• Caveat: constants are absolute prices. They are invalidated by splits, contract rollover in futures and by the simple passage of time as the market moves away.

(b) Indicator plot

References an individual output line of an indicator. You select the indicator, set its parameters, then select which of its plots to compare.

• Example: add RSI, set period to 14, select the RSI plot, operator >, value 60.
• Multi-plot indicators: an indicator such as Volume Profile exposes several plots (POC, value area high, value area low). Each plot is separately selectable, which is exactly how level-based alerts are built.
• Plot-versus-plot comparisons are possible: for example a fast moving average plot crossing above a slow moving average plot.
• Caveat: the indicator parameters inside the alert are independent of the parameters on your chart. If your chart shows RSI(9) and the alert says RSI(14), they will disagree. Match them intentionally.

(c) Order-flow event

Reads metrics computed from the trade-by-trade tape rather than from closing price. Available metrics include bar delta, cumulative delta, buy volume, sell volume and imbalance measures — the same family of data that powers the Footprint chart.

• Example: Delta > 800 on a 30-second bar — fires when net aggressive buying in that half-minute exceeds 800 contracts or shares.
• Negative thresholds capture the sell side: Delta < -800 is aggressive selling.
• Caveat: order-flow thresholds are instrument-specific and interval-specific. A delta of 800 is a violent burst on MES over 30 seconds and unremarkable on ES over 15 minutes. Never copy a threshold between instruments without recalibrating.
• Data requirement: the connection must supply tick data with aggressor side. A trades-only or delayed aggregated feed cannot compute delta.

(d) Session / series data

Compares values from defined session segments, so you can express relationships between the regular session and the overnight session.

• Example: RTH close > overnight close — a directional bias filter stating that the day session finished above where the overnight session finished.
• Other useful comparisons: current price versus overnight high, current price versus prior RTH low, current session VWAP versus prior session close.
• Caveat: session boundaries depend on the instrument's session template and your platform time zone. Verify that the overnight window is what you expect (for US index futures, typically 18:00–09:29 ET) before relying on a session comparison.

(e) Volume

Compares the volume of the current bar against a constant or against the volume of the prior N bars.

• Expansion filter: Volume > volume of prior 2 bars — participation is increasing, often used to confirm a break.
• Contraction filter: the same condition wrapped in NOT — Volume NOT greater than prior 2 bars — participation is drying up, used to detect coiling before expansion or a failing move.
• Caveat in forex: FX volume is tick volume (a count of price updates) or broker fill volume, not consolidated exchange volume, so absolute thresholds are not comparable across brokers.

Combining Conditions: Any/All, Lookback Bars, Groups and NOT

A single condition row is rarely enough for a useful alert. SabrTrader combines rows with Boolean logic, a bars lookback window, and nesting.

The Any / All match toggle

At the top of the condition area is a match checkbox that controls how the rows combine.

Toggle state	Logic	Fires when	Effect on alert frequency
Checked (Any)	OR	At least one condition is true	More alerts, broader coverage, more false positives
Unchecked (All)	AND	Every condition is true	Fewer alerts, higher selectivity, risk of never firing

Getting this backwards is the single most common alert configuration error. If you built a two-condition confluence rule and it fires constantly, check whether Any is still selected.

The "within N bars ago" lookback window

Under All (AND) logic, requiring every condition to be true on the exact same bar is extremely restrictive — real markets rarely align perfectly on one bar. The lookback window relaxes this: it defines how many recent bars a condition may have been satisfied in and still count toward the combined trigger.

• Lookback = 0 (same bar): all conditions must be true simultaneously on the current bar.
• Lookback = 1: a condition satisfied on the previous bar still counts.
• Lookback = 2: a condition satisfied up to two bars ago still counts — this is the setting used in the demonstration.

Practical reading of a two-condition alert with lookback 2: "RSI(14) went above 60 and delta went above 800, in either order, within a three-bar span." This is how confluence actually appears on a chart: momentum shifts, then aggression confirms, or the reverse.

Wider lookbacks increase hit rate but weaken the causal link between the conditions. Beyond about five bars on an intraday interval, the two events are usually unrelated.

Condition grouping (nesting)

Grouping lets you evaluate a set of conditions together, then combine that result with other conditions. This is required whenever your rule mixes AND and OR.

Without grouping, A OR B AND C is ambiguous. With grouping it is explicit:

• (A AND B) AND C — two conditions must both be true, and a third must also be true.
• (A AND B) OR (C AND D) — either of two distinct setups triggers the alert. This is how you monitor a long setup and a short setup with one alert.
• (A OR B) AND C — either of two entry triggers, but always filtered by a mandatory condition C such as a session or volume filter.

The NOT operator

NOT inverts a single condition or a group. It is how you express absence, which is often the real filter.

• NOT (Volume > volume of prior 2 bars) — volume is contracting. Useful to detect quiet consolidation rather than an already-expanding move.
• NOT (Price > overnight high) — price has not yet taken out the overnight high, keeping the alert focused on pre-breakout conditions.
• NOT (Delta > 0) — the bar is net sell-aggressive.

Truth table reference

A	B	A AND B	A OR B	NOT A	A AND NOT B
true	true	true	true	false	false
true	false	false	true	false	true
false	true	false	true	true	false
false	false	false	false	true	false

Three worked composite examples

Intent	Configuration
Momentum confirmed by aggression	Symbol AAPL, 5-minute. Condition 1: Indicator plot, RSI(14) > 60. Condition 2: Order-flow event, delta > 800. Match: All (unchecked). Lookback: 2 bars.
Day session strength taking out a profile level	Symbol MES, 2000-tick. Condition 1: Session data, RTH close > overnight close. Condition 2: Price level, close crosses above value area high. Match: All. Lookback: 1 bar.
Quiet coil at a demand zone	Symbol MES, 30-second. Group 1: (Price low < demand zone high) AND (Price low > demand zone low). Condition 3: NOT (Volume > volume of prior 2 bars). Match: All. Lookback: 2 bars.

Notification Channels

A notification channel is the delivery method for a fired alert. Multiple channels can be attached to a single alert, and channels are configured once and reused across alerts.

Channel	What it does	Best used for
Popup	Displays an in-platform dialog naming the alert	Active trading at the screen; unmissable but interrupts
Sound	Plays an audible tone	Active trading while your eyes are on another chart
Email	Sends a message to the configured address	Away from the desk; also creates a searchable record
Telegram	Pushes a message to a Telegram chat	Mobile monitoring with near-instant delivery
Activity log	Writes a timestamped in-platform entry	Post-session review, journaling and auditing which conditions preceded good trades

Recommended default: attach the activity log to every alert regardless of what else you choose. It costs nothing, never interrupts you, and it is the only way to reconstruct afterwards which alerts fired, when, and whether they were worth acting on.

Selecting no channel at all is legal and produces a silent alert. If an alert appears never to fire, verify that at least one channel is selected.

Trading Concepts Behind the Conditions

Delta

Delta is the net difference between volume traded at the ask (buy aggressors lifting offers) and volume traded at the bid (sell aggressors hitting bids) over a bar or period. Positive delta means aggressive buyers dominated; negative delta means aggressive sellers dominated. A threshold such as delta > 800 asserts that at least 800 contracts of net market-buy pressure was pushed into the tape within one bar — a measure of initiative, not of price movement. Delta is displayed and read in detail on the Footprint chart.

Cumulative delta

Cumulative delta is a running total of bar delta across a session. Where bar delta shows a single burst, cumulative delta shows who has been in control since the session opened. It is useful as an alert filter ("only alert me on longs while cumulative delta is positive") because it changes slowly and describes context rather than a moment.

Order-flow event

An order-flow event is any measurable condition derived from the trade-by-trade tape — delta, buy/sell volume split, imbalances, absorption — rather than from a bar's closing price. Order-flow events describe how price arrived somewhere; price conditions describe only where it arrived.

Absorption

Absorption is aggressive market orders being fully consumed by resting limit liquidity without price making progress. It appears as heavy one-sided volume printing repeatedly at a single price. An alert can approximate it by combining large delta with a small bar range.

Volume profile, POC, value area, HVN and LVN

A volume profile is a histogram of volume traded at each price over a chosen session or range. Its components:

• Point of control (POC) — the single price with the highest traded volume in the profile. The market's consensus price for that period, and a common magnet.
• Value area (VA) — the price range containing a defined percentage of the profile's volume, conventionally 70%, bounded by the value area high (VAH) and value area low (VAL). Acceptance inside the value area suggests balance; movement outside it suggests initiative.
• High volume node (HVN) — a price with unusually heavy volume. Price tends to slow and rotate there.
• Low volume node (LVN) — a price with unusually light volume. Price tends to move through quickly, making LVNs useful as break-confirmation levels.

Full detail is on the Volume Profile indicator page. Each of these values is an indicator plot, which is precisely why the AI can convert them into price-level alerts.

RTH, overnight/ETH and session split profiles

The RTH session is regular trading hours — for US index futures and equities typically 09:30–16:00 ET. The overnight or ETH session is trading outside those hours, commonly 18:00–09:29 ET for index futures. Overnight volume is thinner and driven by fewer participants, so overnight highs, lows and profile levels frequently act as reference points that the day session tests.

Session split profiles means building separate volume profiles for the overnight and RTH segments instead of one blended profile. This matters because a blended profile hides which participants built which level. A POC created by thin overnight rotation behaves differently from a POC created by heavy cash-session participation.

Supply and demand zones; support and resistance

A supply zone is a price area where prior heavy participation suggests resting sellers may react again; a demand zone is the buy-side equivalent. Support and resistance are the broader, price-only version: areas where previous trade has repeatedly halted or reversed movement. Volume-profile-derived zones are stronger than eyeballed lines because they are anchored to where business actually transacted. Structural interpretations of the same idea appear in the Market Structure indicator.

RSI (Relative Strength Index)

RSI is a bounded momentum oscillator, scaled 0–100, comparing the average size of recent gains to the average size of recent losses over a lookback period, commonly 14 bars. Readings above 60 indicate that upward momentum has dominated recently; readings below 40 indicate the reverse. In alerts RSI is best used as a filter (a state you require) rather than a trigger (the event you act on), because it lags the tape.

Indicator plot

An indicator plot is one individual output value or line of an indicator. RSI has a single main plot; a volume profile exposes several (POC, VAH, VAL and more). Only plots can be referenced in alert conditions.

Tick chart

A tick chart completes each bar after a fixed number of trades rather than after a fixed amount of time. A 2000-tick MES chart closes a bar every 2000 transactions. Tick bars normalise for activity: they appear rapidly when the market is busy and slowly when it is quiet, which makes order-flow thresholds more consistent across the day than time bars do.

Entry signal

An entry signal is a rule set identifying a potential trade location. Alert conditions and entry-signal conditions are structurally the same thing; the difference is that an alert notifies and a strategy acts. A well-validated alert is therefore a natural specification to hand to the AlgoStudio Pro designer.

Tutorial A — Build a Simple Price-Break Alert

Goal: notify you when GOOGL closes above 500 on a 15-minute bar, whether or not the chart is open.

1. In the Control Center, choose New > Alerts. The alert manager opens.
2. Click New Alert.
3. In Name, type a descriptive label such as GOOGL level break 500. Naming it just level break works but tells you less when the notification arrives.
4. In Symbol, enter or look up GOOGL. Confirm the resolved instrument and exchange are correct.
5. In Timeframe / interval, select 15 minutes. Conditions will now be evaluated per 15-minute bar.
6. Click Add condition. In the source dropdown choose Price level.
7. Set the series to Close, the operator to greater than (or crosses above if you want a single trigger on the transition bar only), and the value to 500.
8. Leave the match toggle in its default state — with one condition, Any and All are equivalent.
9. Select notification channels. For a simple level alert, popup plus activity log is a reasonable minimum; add Telegram if you will be away from the desk.
10. Click Save. The alert appears in the user-created section of the alert manager and is armed immediately.

Verification: the alert fires without a GOOGL chart open anywhere. To confirm it is live, temporarily set the level just above current price on a fast interval, watch it trigger, then edit it back to the real level.

Tutorial B — Build a Two-Condition Confluence Alert

Goal: notify you when AAPL shows both momentum (RSI above 60) and aggressive buying (delta above 800) on a 5-minute chart, within a two-bar window.

1. Open New > Alerts and click New Alert.
2. Name it AAPL 5m RSI60 + delta800. The name states both conditions so the notification is self-explanatory.
3. Set Symbol to AAPL.
4. Set Timeframe to 5 minutes.
5. Add condition 1. Source: Indicator plot. Choose RSI. Set the period parameter to 14. Select the RSI plot. Operator greater than. Value 60.
6. Add condition 2. Source: Order-flow event. Choose Delta. Operator greater than. Value 800. Note that this value must be calibrated for AAPL on 5-minute bars — inspect recent bars on a Footprint chart to find what an unusually large delta actually is for this instrument before committing to a number.
7. Leave the Any checkbox unchecked so the match logic is All (AND). Both conditions must now be satisfied.
8. Set the within N bars ago lookback to 2. The conditions may now be satisfied on different bars, in either order, inside a three-bar span.
9. Select channels: popup for immediate visibility, Telegram for mobile, activity log for the audit trail.
10. Click Save.
11. Test and verify. Let the alert run through an active session. Open the activity log and inspect the timestamps of each trigger. For each entry, load an AAPL 5-minute chart at that timestamp and check whether the setup was actually worth looking at. This is how you calibrate the delta threshold: if you get twenty triggers a day, raise it; if you get none in a week, lower it or widen the lookback.

Variations worth building from this template

• Short side: RSI(14) < 40 AND delta < -800.
• Add a session filter: group the two conditions and require that the current time is within RTH, eliminating thin overnight triggers.
• Add a volume expansion filter: a third condition requiring volume greater than the prior two bars, so you only get alerted when participation confirms.

Tutorial C — AI: Profile the Sessions, Then Auto-Create Level Alerts

Goal: have the AI assistant apply Volume Profile to a 2000-tick MES chart, split overnight from RTH, mark supply and demand zones, and then generate price-level alerts on the resulting levels.

Prerequisite: the AI assistant must already be configured in SabrTrader, and the chart you want to modify must be the focused chart.

1. Open the chart you want the AI to work on. In the demonstration this is a 2000-tick MES chart with roughly seven days of history loaded. Loaded history matters — the AI can only profile the sessions that are actually present on the chart.
2. Open the AI chat panel.
3. Paste a specific, saved prompt. The one used in the demonstration is:
   Open the volume profile indicator, apply it to this chart, separate the overnight from the RTH session, draw a volume profile for both, and mark the important supply and demand zones on the chart.
   Each clause is a discrete instruction: which indicator, which chart, how to segment sessions, how many profiles, what to annotate. Vague prompts produce vague results.
4. Press Send and wait. Multi-step chart work takes time.
5. Read the AI's written plan before doing anything else. The response describes what it intends to change, including the session windows it inferred. Confirm the overnight window is what you expect — 18:00 to 09:29 for US index futures — and that RTH is defined correctly.
6. Press Apply. This is the confirmation step: until Apply is pressed, nothing has been committed to the chart.
7. Verify the result visually. You should see a separate overnight profile and a separate RTH profile for each loaded day, plus marked supply and demand zones on the current session.
8. Send a second prompt: Create alerts for important levels. The AI reads the levels produced by the indicators it just applied and proposes alerts on them.
9. Press Apply again to commit the alerts.
10. Open New > Alerts. The manager now shows two groups: the alerts you created yourself, and the AI-generated alerts.
11. Select one AI alert and click Edit. Confirm that it is an ordinary price-level condition — for example Close crosses above 77.61 — where the constant corresponds to a profile level such as a POC, VAH or VAL.
12. Attach or adjust notification channels on the AI alerts. Generated alerts may not have the channels you prefer, so review them before relying on them.

Making this a daily routine

Save the prompt text in a plain text file or the platform's saved-prompt library. Before the open, run it across each market you follow, then run create alerts for important levels. You now have level alerts on every market on your list without having drawn a single line by hand. When something triggers, you open that chart and decide whether it is actionable.

Working With AI-Generated Alerts

AI-generated alerts are not a separate class of object. Once created they are ordinary SabrTrader alerts with every normal capability.

• Editable. Open one and you will find a plain price-level condition, for example Close crosses above 77.61. You can change the operator, adjust the constant, change the series from close to high, or add further conditions to turn it into a confluence alert.
• Renameable. Rewrite generic names into meaningful ones such as MES overnight VAH 77.61 so the notification tells you which level was crossed.
• Channel-configurable. Attach email, Telegram, sound, popup or activity log exactly as with a manual alert.
• Deletable and disableable individually or as a group.

They are snapshots, not live links

This is the most important operational fact. An AI-generated level alert stores a fixed price constant that was correct at the moment of generation. It is not bound to the indicator plot. Consequences:

• When the session rolls, a new overnight profile forms with a new POC, VAH and VAL. Yesterday's alerts still reference yesterday's prices.
• If you change the profile's settings, lookback or session definition, existing alerts do not move.
• In futures, contract rollover changes the price series; old constants may be meaningfully offset.

The remedy is a daily routine: delete or disable the previous day's AI alert group, re-run the prompt pre-open, and generate a fresh set. If you need a condition that tracks a level dynamically, build it manually as an indicator plot condition (close crosses above the Volume Profile VAH plot) rather than as a price constant.

Not limited to volume profile

Volume profile is a convenient use case for day traders because it produces clean, meaningful horizontal levels. The same workflow applies to any indicator the AI can access — moving averages, pivot levels, market structure levels, price-action swing levels from the TDU Price Action indicator, or any of the premium indicators. Name the indicator explicitly in the prompt and the AI will read its levels the same way.

Best Practices

Configuration

• Match interval to condition type. Order-flow conditions belong on fast intervals (30-second, tick bars). Momentum and structure filters belong on the interval you actually trade.
• Calibrate thresholds per instrument. Before setting a delta threshold, look at recent bars on a footprint chart for that exact symbol and interval and choose a value that occurs a handful of times per session, not dozens.
• Prefer "crosses above/below" to "greater/less than" for level breaks, so the alert fires once on the transition rather than on every bar while price remains beyond the level.
• Require confluence. Two independent conditions cut noise dramatically compared with one. Momentum plus order flow, or level plus volume, are the most useful pairings.
• Use the bars lookback rather than demanding same-bar coincidence. A lookback of 1–2 bars reflects how confluence really appears without letting unrelated events combine.
• Name alerts so the notification is self-explanatory. Include symbol, interval and the essence of the condition.
• Always attach the activity log in addition to whatever interrupt channel you choose.

Operations

• Keep a small number of high-quality alerts. Five alerts you trust beat forty you ignore.
• Use background alerts for markets you do not watch. This is the highest-value use of the system — it extends your coverage without extending your screen space.
• Regenerate AI level alerts every morning as part of your pre-open routine, and clear out the stale group first.
• Audit with the activity log. Weekly, review every trigger and ask whether it preceded a situation you would want to trade. Delete or tighten the conditions that consistently did not.
• Keep a reusable prompt library. Store prompts in a text file with a short note on what each one produces, so pre-market setup is copy-paste rather than composition.
• Always read the AI's plan before pressing Apply. Apply is the point of no return for chart changes; the plan text is where you catch a wrong session window or wrong indicator.
• Graduate proven logic to a strategy. When an alert's conditions reliably identify locations you would trade, rebuild them in AlgoStudio Pro where they can be backtested, sized and automated. Validate the same logic against recorded tape in Market Replay first.

Common Mistakes

Mistake	Why it fails	Correct approach
Copying a delta threshold from another instrument or interval	Delta scales with contract size, liquidity and bar duration. 800 on MES 30-second and 800 on ES 15-minute describe entirely different events.	Measure typical and extreme delta on the specific symbol and interval before choosing a number.
Leaving Any checked when you meant All	Each condition fires the alert independently, so a carefully built confluence rule degenerates into two noisy single-condition alerts.	Uncheck the match box for AND logic. Verify by counting triggers on the first session.
Demanding all conditions on the same bar	Lookback 0 is so restrictive the alert may never fire, and the trader concludes the feature is broken.	Set the lookback to 1–2 bars.
Expecting AI level alerts to update themselves	They store fixed price constants captured at generation time, not live references to the indicator plot.	Regenerate them pre-open, or build the condition manually against the indicator plot.
Alert stacking / alert fatigue	Dozens of loosely filtered alerts produce constant notifications, and the trader learns to dismiss them without looking — including the good ones.	Cap the number of active alerts. Delete anything that has not produced a useful look in two weeks.
Treating an alert as a trade signal	An alert has no risk model, no position sizing and no exit. Acting on it mechanically is trading a rule you never tested.	Treat every alert as a "look here" prompt. If you want mechanical execution, build a strategy.
Forgetting to select a notification channel	The alert evaluates and fires silently, so it appears not to work.	Confirm at least one channel is selected; make activity log a habit.
Setting alerts on an interval you never trade	A 1-minute alert delivered to a trader who executes on 15-minute charts arrives too early and too often to be usable.	Align the alert interval with your decision interval, using fast intervals only for order-flow conditions inside a confluence rule.
Assuming the chart must be open	Traders keep unnecessary charts loaded, consuming market data lines and screen space.	Alerts run independently of the workspace. Close the chart; the alert still fires.
Vague AI prompts	"Add some profiles" leaves session windows, profile count and annotation entirely to inference.	Name the indicator, the chart, the exact session windows and the annotations you want.
Setting order-flow conditions on a delayed or aggregated feed	Delta requires trade-by-trade data with aggressor side; without it the condition cannot evaluate.	Verify your connection supplies real-time tick data for that symbol.

Frequently Asked Questions

Does the chart or symbol need to be open in my workspace for an alert to fire?

No. Alerts are evaluated by the platform independently of your layout. The symbol does not need a chart open, and the alert does not need to belong to the workspace you are currently viewing. This is what makes alerts the correct tool for monitoring markets you are not actively watching. The only requirement is that a connection capable of supplying the necessary data for that symbol is active — and for order-flow conditions, that connection must deliver trade-by-trade tick data.

What is the difference between the Any (OR) and All (AND) matching options?

Any (checkbox checked) is OR logic: the alert fires when at least one condition is satisfied. All (checkbox unchecked) is AND logic: the alert fires only when every condition is satisfied. Any produces more alerts and broader coverage with more false positives; All produces fewer, more selective alerts and risks never firing if the conditions are too strict. If you built a confluence rule and it is firing constantly, the most likely cause is that Any is still selected.

How does the 'within N bars' lookback change when an alert triggers?

The lookback window defines how many recent bars a condition may have been satisfied in and still count toward the combined trigger. With a lookback of 0, all conditions must be true on the same bar. With a lookback of 2, a condition satisfied up to two bars ago still counts, so the conditions may occur on different bars in either order within a three-bar span. Wider lookbacks increase how often the alert fires but weaken the connection between the events — beyond roughly five bars on an intraday interval the conditions are usually unrelated.

Can I create an alert on delta or other order-flow data?

Yes. Select Order-flow event as the condition source. Available metrics include bar delta, cumulative delta, buy volume, sell volume and imbalance measures — the same data family that drives the Footprint chart. A typical condition is Delta > 800 on a 30-second bar. Order-flow conditions require a connection that supplies tick-by-tick trades with aggressor side; a delayed or pre-aggregated feed cannot compute delta.

What delta threshold should I use, and why does the bar interval matter?

There is no universal number. Delta scales with the instrument's contract size and liquidity, and with how much time or activity the bar aggregates. A delta of 800 in a 30-second MES bar is a violent burst of aggressive buying; the same 800 accumulated over a 15-minute bar is routine. To calibrate, put a footprint or delta display on the exact symbol and interval you intend to alert on, observe values over a full session, and pick a threshold that occurs a handful of times per day rather than dozens. Never copy a threshold between instruments or intervals without re-measuring.

How do I send alerts to Telegram or email?

Both are notification channels selected in the alert dialog. Configure the channel credentials once in platform settings — the destination email address, or the Telegram chat the platform should push to — and thereafter simply tick the channel on each alert you create. Multiple channels can be attached to a single alert, so one alert can produce a popup at your desk and a Telegram push on your phone simultaneously. If messages do not arrive, verify the channel credentials before changing the alert conditions.

How do I ask the AI to create alerts from volume profile levels?

Two prompts. First apply the indicator and levels, for example: Open the volume profile indicator, apply it to this chart, separate the overnight from the RTH session, draw a volume profile for both, and mark the important supply and demand zones. Read the AI's plan, then press Apply. Second, send Create alerts for important levels and press Apply again. The AI reads the levels produced by the applied indicator and creates price-level alerts on them. Open New > Alerts to see them in the AI alerts group.

Do AI-generated alerts update automatically when the session rolls or levels change?

No. Each AI-generated level alert stores a fixed price constant that was correct at the moment of generation — it is a snapshot, not a live link to the indicator plot. When a new session builds a new POC, value area high or value area low, the existing alerts still reference the old prices. Regenerate them as part of your pre-open routine after clearing the previous day's group. If you need a condition that tracks a level dynamically, build it manually as an indicator-plot condition rather than a price constant.

Can I edit or delete an alert the AI created?

Yes. AI-generated alerts are ordinary alerts. Select one in the alert manager and click Edit to see its condition — typically a plain price-level rule such as Close crosses above 77.61. You can rename it, change the level, change the operator or series, add further conditions to turn it into a confluence alert, attach different notification channels, disable it, or delete it.

Can the AI build alerts from indicators other than volume profile?

Yes. Volume profile is a convenient demonstration because it produces clean horizontal levels that day traders care about, but the workflow applies to any indicator the AI can access — moving averages, pivots, market structure levels, price-action swing levels and others. Name the indicator explicitly in your prompt, have the AI apply it, then ask it to create alerts on the resulting levels.

How is the alert builder different from AlgoStudio Pro or the Strategy Builder?

The alert builder produces notifications; a strategy produces orders. The alert builder can express rich Boolean logic over indicators, order flow, session data and volume, which brings it close to the entry-signal portion of an automated strategy — but it has no position sizing, no exit logic, no trade management and no backtest. AlgoStudio Pro and the Auto Strategy Builder add all of those. A practical progression is to prototype the trigger as an alert, audit its history in the activity log, and then rebuild the validated logic as a strategy.

Can one alert combine an indicator condition, an order-flow condition and a session comparison?

Yes. There is no restriction on mixing source types within a single alert. A valid rule is: RSI(14) greater than 60, AND delta greater than 800, AND RTH close greater than overnight close — all with All matching and a bars lookback of 2. Use condition grouping when you need mixed AND and OR logic, for example (RSI > 60 AND delta > 800) OR (RSI < 40 AND delta < -800) to cover both directions with one alert.

How do I use a NOT condition, for example volume not greater than the prior two bars?

Build the positive condition first — Volume source, operator greater than, compared against the volume of the prior 2 bars — then apply the NOT operator to that row. The result reads "volume is not greater than the prior two bars", which is a volume-contraction filter: it is true when participation is drying up. Combined with a price condition it detects quiet coiling at a level rather than an already-expanding move. NOT can also be applied to a group, inverting the entire group's result.

Why did my AI prompt not change anything on the chart?

The most common reason is that Apply was never pressed. The AI produces a written plan and asks for confirmation; nothing reaches the chart until you confirm. Other causes: the AI assistant is not yet configured; the chart you intended to modify was not the focused chart; the indicator you named is not available on your plan; or the chart does not have enough loaded history for the sessions you asked it to profile.

How many alerts should I run at once without creating alert fatigue?

There is no hard limit in the platform, but there is a practical one for human attention. Alert fatigue is the state in which so many loosely filtered alerts fire that you stop looking at them, which makes the whole system worthless. A workable discipline is a small set of hand-built confluence alerts you genuinely trust — often five to ten — plus a regenerated daily group of level alerts across the markets you follow. Review the activity log weekly and delete or tighten anything that has not produced a useful look.

What is the difference between a price alert and a conditional alert?

A price alert is a single-condition rule on a fixed price level: it fires when price trades or closes above or below that level. A conditional alert has one or more logical conditions on price, indicators, order flow, session data or volume, combined with AND, OR and NOT operators and an optional bars lookback. Every price alert is technically a conditional alert with one price-level condition; the term "conditional" is used for the multi-condition case.

Does the alert fire on bar close or intrabar?

It depends on the series you compare. A condition written against close can only be judged when the bar completes, so the alert fires at bar close. Conditions written against last price, high, low, or using crosses above/below can become true while the bar is still forming and fire intrabar. Bar-close conditions are later but more reliable; intrabar conditions are earlier but can un-happen before the bar completes.

Why should overnight and RTH volume profiles be split rather than blended?

Because a blended profile hides which participants built which level. The overnight or ETH session is thinner and driven by fewer participants, so a point of control formed overnight represents much less committed business than one formed during regular hours. Splitting the sessions lets you read each set of levels on its own terms and lets you compare them — for example, alerting when the RTH session closes above the overnight close. Details of profile construction are on the Volume Profile indicator page.

Why use a tick chart such as 2000-tick MES for order-flow alerts?

A tick chart completes each bar after a fixed number of trades rather than after a fixed amount of time. This normalises the bar for activity: bars form quickly when the market is busy and slowly when it is quiet. As a result, an order-flow threshold such as a delta level stays comparably meaningful across the open, mid-session lull and close, whereas on a time chart the same threshold means very different things at 09:30 and 13:00.

Can I use an alert as an automated entry signal?

An alert can define the same conditions an entry signal would, but it does not place orders. Treat an alert as a prompt to look at a chart and make a decision. If you want mechanical execution, rebuild the validated conditions in AlgoStudio Pro, where they can be paired with position sizing, exits and risk management, and tested against history before deployment.

Troubleshooting

The alert never fires

Check in this order:

• Is the alert enabled? Disabled alerts remain in the list and are silently inert.
• Is a notification channel selected? With no channel, the alert evaluates and fires but delivers nothing observable.
• Is the match logic All when it should be Any? AND logic across several strict conditions can be effectively impossible.
• Are the conditions mutually exclusive? For example "close greater than 500" AND "close less than 490" can never both be true.
• Is the lookback 0? Requiring all conditions on the same bar is very restrictive. Try 1 or 2 bars.
• Is the threshold unreachable? A delta threshold set far above anything the instrument produces on that interval will never trigger. Measure real values first.
• Is the interval too slow? A 15-minute alert can only fire four times an hour at most.
• Is the market open? No ticks means no bar completion and no evaluation.
• Does the feed supply the required data? Order-flow conditions need real-time tick data with aggressor side.

The alert fires constantly

• Check the Any/All toggle. If Any is checked, each condition triggers independently. Uncheck it for AND logic.
• Change "greater than" to "crosses above". A greater-than comparison stays true on every bar while price remains beyond the level and can re-fire indefinitely.
• Raise the threshold. A delta or volume threshold set near the instrument's typical value will trigger on ordinary bars.
• Add a filter condition. A volume expansion requirement, a session restriction, or a cumulative delta direction filter will remove most low-quality triggers.
• Narrow the lookback. A wide lookback allows unrelated events to combine.
• Move to a slower interval if the alert is a structural or momentum rule rather than an order-flow trigger.

Telegram or email notifications are not arriving

• Confirm the channel is actually ticked on the specific alert — channels are per-alert, not global.
• Verify the channel credentials in platform settings: the destination email address, or the Telegram chat and bot configuration.
• Check the activity log. If the log shows the alert fired but no message arrived, the problem is in the channel configuration or delivery, not the alert logic.
• Check spam and promotions folders for email, and confirm the Telegram chat is not muted.
• Test with a deliberately easy condition — a price level just above current price on a fast interval — so you get an immediate trigger to diagnose against.

The AI did not apply anything to the chart

• Apply was not pressed. The AI produces a plan and requests confirmation; nothing is committed until you press Apply. This is the most common cause.
• The AI assistant is not configured. Complete AI setup before using the chat panel.
• The wrong chart was focused. The AI acts on the chart in focus. Click the target chart before sending the prompt.
• The named indicator is unavailable. Confirm the indicator exists on your plan — see Free Indicators and Premium Indicators.
• Insufficient loaded history. The AI can only build profiles for sessions present on the chart. Increase the history lookback and retry.

AI alerts do not appear in the alert manager

• Confirm you pressed Apply after the second prompt ("create alerts for important levels"). The plan alone creates nothing.
• Confirm the indicator levels existed when you asked. If the profiles were never applied, there are no levels to alert on.
• Reopen New > Alerts and look specifically in the AI alerts group — it is displayed separately from your hand-built alerts and is easy to overlook.
• Confirm the correct chart and workspace were focused when the prompt was sent.

The AI split the sessions at the wrong times

The AI infers session windows unless you specify them. Restate the prompt with explicit times, for example: Treat 18:00 to 09:29 as the overnight session and 09:30 to 16:00 as the RTH session, and build a separate volume profile for each. Also verify your platform time zone, since session boundaries are interpreted against it. Read the AI's written plan before pressing Apply — the inferred windows are stated there, which is where you catch the error before it reaches the chart.

Level alerts are stale after the session rolled

This is expected behaviour, not a fault. AI-generated level alerts store fixed price constants captured at generation time; they do not follow the indicator. Build a pre-open routine: delete or disable the previous AI alert group, re-run the profiling prompt, then re-run the alert-creation prompt. For futures, also regenerate after contract rollover, since the price series shifts. If you want a level that tracks dynamically, build the alert manually with an indicator plot condition referencing the profile plot instead of a constant.

The indicator value in the alert does not match what I see on my chart

Indicator parameters inside an alert are independent of the parameters on any chart. If your chart displays RSI with a period of 9 and the alert condition uses RSI with a period of 14, the two will disagree. Open the alert, inspect the indicator parameters in the condition row, and set them to match the chart you make decisions on. The same applies to session settings and lookbacks on profile-type indicators.

An order-flow condition cannot be selected or returns no data

Order-flow metrics are computed from trade-by-trade data including aggressor side. If your connection for that symbol supplies delayed data, end-of-day data or pre-aggregated bars only, delta and related metrics cannot be calculated. Verify the data entitlement for that asset class on your connection, and confirm the symbol is one your feed carries tick history for. See the relevant connection guide, for example Connecting SabrTrader to Rithmic, for entitlement requirements.

I want two directions in one alert but it only catches longs

Use condition grouping. Build the long conditions as one group and the short conditions as another, then combine the two groups with Any (OR) logic: (RSI > 60 AND delta > 800) OR (RSI < 40 AND delta < -800). Without grouping, mixing AND and OR at a single level is ambiguous and will not express what you intend. Include the direction in the alert name so the notification tells you which side triggered.

Glossary

Absorption

Aggressive market orders being fully consumed by resting limit liquidity without price making progress, typically appearing as heavy one-sided volume printing repeatedly at a single price.

Activity log

A timestamped in-platform record of alert triggers, used for post-session review, journaling and auditing which conditions preceded good trades.

AI apply/confirm workflow

The two-step process in which the AI assistant proposes chart or alert changes in a written plan and nothing is committed until you press Apply.

Alert fatigue / false positives

The condition in which too many loosely filtered alerts fire, training the trader to ignore notifications — including the useful ones.

Bar close vs intrabar trigger

Whether a condition is evaluated only when a bar completes (bar close) or in real time as the bar forms (intrabar). Bar-close conditions fire later but are more reliable; intrabar conditions fire earlier but can un-happen.

Boolean logic (AND / OR / NOT)

Operators that combine alert conditions so that all conditions (AND), any condition (OR), or the absence of a condition (NOT) triggers the alert.

Condition grouping

Nesting condition sets so that each group is evaluated together before its result is combined with other conditions or groups. Required whenever a rule mixes AND and OR logic.

Conditional alert

An alert that fires only when one or more logical conditions on price, indicators, order flow, session data or volume are satisfied, combined with AND, OR and NOT operators.

Cumulative delta

A running total of bar delta across a session, showing whether buyers or sellers have been in control since the session opened.

Delta

The net difference between volume traded at the ask (buy aggressors) and volume traded at the bid (sell aggressors) over a bar, session or range. Positive delta indicates aggressive buyers dominated; negative delta indicates aggressive sellers dominated.

Entry signal

A rule set that identifies a potential trade location. The same logic that powers an alert can seed an automated strategy, but an entry signal in a strategy also carries sizing, exits and risk management.

Footprint chart

A chart type that displays traded volume split by buy and sell aggressor at each individual price level within each bar, used to read absorption, imbalance and delta. It requires trade-by-trade tick data.

High volume node / low volume node

Profile prices with unusually heavy volume (HVN) or unusually light volume (LVN). HVNs tend to attract and hold price; LVNs tend to be traversed quickly.

Indicator plot

An individual output line or value of an indicator that can be referenced in an alert condition. Multi-output indicators such as volume profile expose several plots, including POC, VAH and VAL.

Lookback window

The number of recent bars in which separate alert conditions may occur and still count as a combined trigger, configured as 'within N bars ago'.

MES

The Micro E-mini S&P 500 futures contract, one tenth the notional size of ES; a liquid, low-notional instrument commonly used for learning a platform and practising order-flow reading.

Order-flow event

A measurable condition derived from the trade-by-trade tape — delta, buy/sell volume split, imbalances, absorption — rather than from a bar's closing price.

Overnight / ETH session

Trading outside regular hours, commonly 18:00–09:29 ET for US index futures. Volume is thinner and driven by fewer participants, and its profile levels often act as reference points for the following day session.

Point of control (POC)

The single price level in a volume profile with the highest traded volume; the market's consensus price for that period.

Price alert (level break)

An alert that fires when price trades or closes above or below a fixed price level.

Prompt engineering

Writing a clear, specific instruction so the AI assistant applies the exact indicators, session windows, levels and annotations you want, rather than inferring them.

RSI (Relative Strength Index)

A bounded momentum oscillator scaled 0–100 that compares the average size of recent gains to the average size of recent losses over a lookback period, commonly 14 bars.

RTH session

The regular trading hours segment of the day, typically 09:30–16:00 ET for US index futures and equities, when participation is highest.

Session split profiles

Building separate volume profiles for the overnight and RTH segments instead of one blended profile, so each session's levels can be read on their own terms.

Supply and demand zone

A price area where prior heavy participation suggests resting sellers (supply) or buyers (demand) may react again.

Support and resistance

Price areas where previous trade has repeatedly halted or reversed movement.

Telegram notification

A push message delivered to a configured Telegram chat when an alert fires, used for monitoring away from the platform.

Tick chart

A chart on which each bar completes after a fixed number of trades rather than after a fixed amount of time, normalising bars for market activity.

Timeframe / interval

The bar aggregation — time, tick, volume or range — on which an alert's conditions are evaluated. It determines what any threshold in the alert actually means.

Value area (VAH/VAL)

The price range containing a defined percentage of a profile's volume, conventionally 70%, bounded above by the value area high (VAH) and below by the value area low (VAL).

Volume profile

A histogram of volume traded at each price over a chosen session or range, used to locate high-activity price areas such as the point of control and value area.

Watchlist monitoring

Using background alerts to track markets that are not displayed on screen, extending coverage without consuming layout space.

Related Documentation