Cival Trapped Traders — Platform Compatibility

The ATAS plugin is the reference implementation. Each port computes what its platform's data allows and labels anything approximate or unavailable inside the indicator itself. The sections below are the READMEs that ship next to each port's source file (ports/<platform>/README.md).

TradingView (Pine Script v6)

Cival Trapped Traders for TradingView (Pine Script v6)

Port of the ATAS original Cival Trapped Traders v1.2.1. It maps closed-bar failed-aggression hypotheses: directional executed volume outside a known structural reference (source, "at risk"), a later close back through that reference ("reclaimed"), a still-later qualifying displacement ("confirmed"), and the observed aftermath (rejected retest, invalidation, expiry). It never infers positions, identities, stops or exits, and the 0-100 evidence points are a checklist, not a probability. Nothing here trades or sends orders.

Requirements

• TradingView Premium or Ultimate plan. The detector is built on request.footprint(), which returns na on lower plans. This mirrors ATAS, where the source rule makes footprint mandatory: without footprint data the source detector is OFF and the dashboard says so. There is no price-only or lower-timeframe substitute for the source rule, because per-price-row aggression, diagonal imbalances and candle-extreme rows cannot be reconstructed from bars.
• A symbol and timeframe for which TradingView can build a volume footprint (history depth varies by plan and timeframe). Use standard candles/bars.
• File: CivalTrappedTraders.pine, overlay indicator, Pine v6.

Install

1. Open the Pine Editor on TradingView, choose Open > New blank indicator, paste the contents of CivalTrappedTraders.pine, click Save, then Add to chart.
2. Leave Apply a starting preset on Custom (identical to Balanced at default values) or pick Balanced / Selective / Responsive.
3. Alerts are off by default. To use them: turn on Enable new live local alerts, then create a TradingView alert on this indicator with condition Any alert() function call. Alerts fire only on live bar closes, never for history.
4. If TradingView reports a runtime/time-limit error on instruments with thousands of ticks per bar, raise Footprint ticks per row (group 98 TradingView data) or lower Maximum chart slots at rebuild.

Feature matrix

Values: Full / Approximate / Not available / Not ported.

Settings mapping

Input names, groups and defaults are the ATAS ones (see the shared settings reference). Present in this port: 00 Quick setup, 00 Chart clarity, 01 Structure, 02 Source aggression, 03 Prior baselines, 04 Later confirmation, 05 Zone lifecycle, 70 Review (Review as of bar), 80 Data and limits, 89 Live visuals (extend boxes, arrows, confluence), 90 Display (all but the items below), 91 Colors, 95 Alerts (not the sound).

TradingView-only input: 98 TradingView data > Footprint ticks per row (default 1).

Not ported

Known differences from ATAS results

1. Row prices. ATAS reads one bid/ask row per traded price. TradingView returns rows of ticks per row ticks with an upper and lower edge. The script uses each row's lowest tradable tick as its price (down_price + max(0, up_price - down_price - mintick) / 2), which equals the ATAS price at 1 tick per row. This convention is an assumption and has not been checked against live TradingView data. At more than 1 tick per row, stacks and imbalances compare rows of that size, so results diverge from ATAS.
2. Footprint availability. TradingView footprint history depth, symbols covered and chart-volume agreement differ from an ATAS feed. Bars without reconcilable footprint count as "unavailable"; sources cannot form on them.
3. Visible zones. ATAS limits to cycles in the scrolled viewport; Pine cannot read the viewport, so "Maximum recent visible zones" keeps the newest N overall.
4. Review mode uses TradingView bar_index, not ATAS ordinals, so the same number points at a different bar.
5. Drawing caps. Boxes (460 budget), lines and labels are capped by TradingView at 500 each; very busy charts drop the oldest profile bars first.
6. Runtime. Row scanning runs only on bars where a detector can possibly fire, but wide-range bars at 1 tick per row can be slow on TradingView. Raise ticks per row if needed.

Verification status

Compiled with TradingView's Pine compiler (pine_check) on 2026-10-26: compiled: true, 0 errors, 0 warnings. Not run on a live chart or in replay. Footprint row conventions, history depth, the runtime on real symbols, alert delivery and drawing appearance are unverified. No claim of profitability, accuracy or vendor parity is made.

NinjaTrader 8 (NinjaScript)

Cival Trapped Traders - NinjaTrader 8

Port of the ATAS original v1.2.1 (Cival.TrappedTraders). Closed-bar failed-aggression study: directional aggression at a known structural reference, a later reclaim and a later qualifying confirmation, followed through to retest, invalidation, expiry or interruption. "Confirmed" means this study's failed-aggression criteria were met on a closed bar. It does not say anything about traders' actual positions, size, entry price or exits, which the study never sees. Evidence points are a checklist out of 100, not a probability. No orders, no account access, no network access.

Requirements

Install

1. Close any open NinjaScript Editor windows.
2. Copy CivalTrappedTraders.cs to Documents\NinjaTrader 8\bin\Custom\Indicators\.
3. In NinjaTrader choose New -> NinjaScript Editor, open the file, and press F5 to compile. If the compiler reports an error, keep the exact message.
4. Open a chart with an Order Flow+ Volumetric series (Ticks per level 1), then right-click -> Indicators -> add Cival Trapped Traders.
5. Start with the defaults (they match the ATAS "Balanced" preset). Allow at least 21 completed bars of footprint history.

Colours: rose = buyers failed (bearish response), teal = sellers failed (bullish response), gold = at risk / unconfirmed.

What is detected

Same rule set as the ATAS original (see its README for the full text):

• Failed breakout - aggressive volume outside a prior range, confirmed swing, or manual reference after a strict price breakout.
• Edge rejection - directional extreme-row aggression at a known boundary that the source candle already rejects.
• Stack failure - consecutive diagonal imbalances (aggressor volume >= ratio x opposite volume one tick away) that fail to hold their source region.
• Lifecycle - at risk -> reclaimed -> confirmed -> rejected retest, with accepted, filtered, expired, invalidated and interrupted as explicit outcomes. A confirmation can never happen on the source bar.
• Data checks - every source needs reconciled rows (row volume within tolerance of candle volume, classified bid+ask >= the configured coverage). Missing or inconsistent rows make the footprint "unavailable" for that bar; nothing is filled in.

Feature matrix

Values: Full = same rule set and thresholds; Approximate = same idea, platform-specific difference described; Not available = the platform cannot supply it; Not ported = not implemented.

Settings mapping

All 77 ported inputs keep the ATAS display names, group names and defaults (taken from the compiled settings dump). Groups: 00 Chart clarity, 00 Quick setup, 01 Structure, 02 Source aggression, 03 Prior baselines, 04 Later confirmation, 05 Zone lifecycle, 70 Review, 80 Data and limits, 89 Live visuals, 90 Display, 91 Colors, 95 Alerts, 99 Export. Decimal settings are double, colours are NinjaTrader brushes.

Not ported (not present as inputs):

Known differences from ATAS

• Different data. The rules are the same, but the rows come from NinjaTrader Volumetric bars instead of ATAS price levels. Feed, tick history and bid/ask classification differ between platforms, so the same market can produce different rows and therefore different candidates. Bar timestamps are NinjaTrader's (bar close time for time-based bars); ATAS used bar open time. Exports record the NinjaTrader timestamp and make no time-zone conversion.
• Arithmetic. The ATAS engine used decimal; this port uses double with a one-thousandth-of-a-tick tolerance on price comparisons so grid-exact ties behave the same.
• Volumetric delta type. With delta type "Up/Down tick" instead of "Bid/Ask", NinjaTrader classifies volume differently and results will differ from a Bid/Ask chart.
• Ticks per level above 1 is refused rather than approximated (see the stack failure row).
• Real-time only alerts. NinjaTrader raises Alert() only in real-time, which matches the original's rule that history never alerts.
• Reloading. Changing any input reloads the indicator and recomputes from the bounded history window, as the ATAS settings changes did.

Verification status

Done offline against stand-ins for the NinjaTrader API, because NinjaTrader is not installed here:

• The engine was run side by side with the ATAS TrapEngine on 96 seeded synthetic footprint scenarios (1,713 zones, 7,131 events, all three source kinds and all lifecycle states). Zones, events, evidence and the 16 output values matched within 1e-7.
• The indicator was driven through load, history and real-time bars against API stand-ins, including exports (JSON parsed successfully).
• The source was syntax- and type-checked as C# 7.3 against .NET Framework reference assemblies.

Not verified: compilation against the real NinjaTrader assemblies, behaviour on a live chart, Volumetric API behaviour on real data, and appearance of the drawings. The scenarios are synthetic; they say nothing about market behaviour and this study makes no claim of profitability or accuracy.