Create and manage background indicator alerts
Run persistent completed-bar monitors across exact feeds, instruments and timeframes, and manage their full delivery and recovery lifecycle.
Reviewed 2026-10-24 · Windows & macOS
On this page
View full-size screenshot ↗
View full-size screenshot ↗Decide what must keep running#
Open Background alerts under Automate. Each saved monitor owns one exact feed, symbol, timeframe, indicator definition/output and condition. It runs on the AMM server independently of the chart, screener, workspace or window that created it. Closing its chart or desktop window does not pause it. The server and required market-data connection must remain running.
A screener rule has a different lifetime: it is evaluated while that screener runs. Use a background alert when you need monitoring independent of that view. A background alert is also separate from a calendar-event reminder, a broker stop and an automatically trading algorithm.
Create one alert#
- Choose New alert. A chart/screener shortcut can prefill exact instrument and indicator settings.
- In What to watch, choose Instrument across the available feeds and the Timeframe. Confirm the feed/account shown, especially for the same ticker on several brokers.
- In When to alert, choose an indicator and output. Use Edit to set its parameters.
- Choose a supported condition and the count for N closed bars in a row. The count accepts 1–100.
- In How to notify, select In-app, Desktop, Sound, Telegram and/or Discord. Inspect each channel's setup status and use Set up… if needed.
- Choose whether Telegram/Discord should receive a chart image and its permitted bar count. Use Send test to check channel delivery.
- Review the generated/editable name, rule sentence and before-saving checklist. Set Start monitoring after saving as intended and save.
An unproven source requires explicit selection before saving; the chart's execution account is not automatically proof of a separate market-data connection. Subsequent changes to the source chart's indicator or timeframe do not rewrite the saved monitor.
Understand exactly when it fires#
Numeric conditions support Above, Below and Equals for the selected output. Equals is exact and carries a warning; floating numerical series may rarely match an exact threshold. A direction/background condition requires a compatible background output, not an arbitrary coloured histogram.
The monitor evaluates settled bars. Repeated updates of one forming candle do not advance the run. A rule fires once when a qualifying run first reaches the configured length. A false, missing or neutral result ends the run and permits a later run to alert again.
For Nova background, Green is +1, Red is −1 and Neutral is 0. With three bars, Green/Green/Green confirms at the third close; a fourth Green does not repeat. With Either, Green and Red are still different runs; Green/Green/Red is not a three-bar run. Other studies can expose positive/negative background states without using those plot colours. These background confirmations are different from histogram colour, line crossings or a complete Nova trading-system entry rule.
Create a mixed batch#
Choose New alert → Multiple instruments… or Watch several instruments… in the single editor. The three steps are Instruments & timeframes, Rule & notifications, then Review & create.
In the matrix, assign only the required intervals to each instrument. For example, EURUSD at 15m/1h and XAUUSD at 1m/5m creates four targets, not every interval for both symbols. Other… opens the full timeframe selector. Clicking a timeframe heading toggles that column for selected rows, or all rows if none is selected. Rows with no timeframe are left out.
The shared instrument picker can combine exact available forex/CFD, equity, spot-crypto and futures feeds. A listed timeframe is not proof every live feed supports it; review validates each pair. Add a name prefix, set the common rule/delivery policy, then inspect New, Already exist · skipped and Can't be created. Use Open existing, Remove from batch or Open Brokers & feeds to resolve issues.
Matching non-archived definitions, including paused ones, are skipped without changing/enabling them. Archived monitors do not block new creation. Up to 500 non-archived monitors, including paused entries, are retained as active capacity. A review preview accepts up to 2,500 target combinations, but that is not the saved-monitor limit. The server rechecks the complete batch at creation; a validation/capacity/persistence failure prevents partial new children appearing.
Monitor state and history#
Status chips include Monitoring, Warming up, Waiting, Needs attention and Paused; show Archived through More. Group by None, Batch, Symbol or Feed. The table shows the rule, run dots, status and last-fired time. It updates automatically every ten seconds and on alert messages; there is no Refresh button.
Select a monitor for Overview, History and Settings. Overview explains attention causes, current run and the last 12 closed bars, plus the next bar close where known. History filters All, Confirmations, Missed and Failed deliveries. Each channel has a delivery result; use Retry delivery on an eligible failure. A condition confirmation and successful external delivery are separate facts.
Edit, pause, archive and export#
Use row/hover menus for Edit, Pause/Resume, Open chart, Duplicate, Send test and Archive. Open chart reproduces the monitor's exact feed/timeframe/indicator. Two or more checked rows expose bulk Pause, Resume, Edit notifications…, Change timeframe…, Archive and Clear. State changes offer Undo where shown.
Changing the condition/indicator resets its baseline/run. Renaming or changing destinations preserves an already-fired run. A new or resumed monitor establishes a silent baseline. Archive stops monitoring while retaining definition/history and removes it from the 500-monitor capacity count. Archived entries offer Restore or Delete permanently…. More → Export list… writes CSV of shown rows; it is a report, not a replacement for a saved-definition backup.
Understand restart and failure recovery#
Enabled monitors restore after server restart with evaluated-bar/run checkpoints, so already processed runs do not simply fire again. Initial history warms the indicator silently. Confirmations first reached during downtime can be recorded as Missed without retroactive notifications; startup does not produce an old-alert storm. Paused and archived monitors stay stopped.
An interrupted external delivery can have an unknown outcome and is not blindly replayed after restart. Exactly-once notification delivery through an external provider is not guaranteed. Source/indicator unavailability appears as attention/waiting state instead of a fabricated healthy neutral stream.
Use the detail's cause and next step: Open Brokers & feeds, Change data feed… or Pause alert where offered. Restore the exact credential/profile/indicator dependency, then inspect current run and history before resuming. Use Messaging and notification delivery for tokens, permissions and reply policy; use Screener rules when the monitor should depend on an active grid instead.
Continue with related tasks#
Configure screener rules and background monitoring, Manage notifications and Telegram or Discord alerts, Add, configure and interpret chart indicators.