Structure Break Risk

It lives in a dedicated RSI pane and draws onto the price chart above it (via force_overlay), because the read needs both:

• RSI pane — the RSI with momentum-divergence lines drawn on it. When price makes a higher high while RSI makes a lower high (uptrend), or price a lower low while RSI a higher low (downtrend), a line marks the failing momentum. Read against the price above — sharing the same time axis — the divergence is unmistakable: the earliest structural crack. A divergence is price vs. momentum and is only readable where both are visible, which is why it lives here rather than as an abstract line on price. A compact “Div” label carries the RSI values in its tooltip.
• Price chart — the decisive break-level line the trend must hold, a shaded risk zone for the cushion between price and that level, a thin Trend Context EMA line, word event labels (Watch / Break Pressure / Structure Broken / Pressure Faded) and the info label.

The risk magnitude — the same 0–100 composite — is reported in a single compact, mobile-friendly info label. Colour encodes the risk band (neutral → orange → red → dark red), never the trade direction, so it never reads as buy/sell; direction is stated in words.

SBR is symmetric by design: in an uptrend it scores top-break risk, in a downtrend bottom-break risk. The asymmetry comes entirely from the situation (which trend is active), never from a hardcoded long/short. The trend context latches — it holds the last clear direction through transition zones, so risk can build before the trend formally flips (exactly where the break is forming). Until a first direction is established the gauge reads No trend and sits at zero.

The headline idea: risk rises as price approaches the decisive break level and extra evidence stacks up — not only after the break is confirmed. A Reason read names the dominant driver in plain language (Approaching break level / Failed breakout / Structure eroding / Momentum divergence / Break confirmed) so the score explains itself.

Pivots are used as swing-level references only (Location); the break event (a close beyond the opposite swing) is the Trigger. The pivot itself is never the signal.

Features

• Four pre-break evidence sensors across the roles, weighted into one score — never an AND-chain — plus the confirmed break (BOS) as a separate event that is not mixed into the score
• Symmetric top-/bottom-break scoring driven by a latching prevailing-trend context
• RSI pane with 70 / 50 / 30 reference levels and momentum-divergence lines drawn on the RSI (price extends but RSI does not), with a compact “Div” label and RSI values in the tooltip — divergence shown where it is actually readable
• Break-level line drawn on price (the decisive swing the trend must hold, coloured by the current risk band) + a risk zone shading the cushion between price and that level, intensifying as risk rises (shown from Watch upward)
• Thin Trend Context line (EMA) on price showing which trend is at risk — colour stays risk-coded, not buy/sell
• Word event labels placed on the side at risk — Watch / Break Pressure / Structure Broken / Pressure Faded — with cooldown; confirmed breaks require an EMA context flip
• State read — No trend / Quiet / Watch / Break Pressure / Critical / Structure Broken — plus a plain-language Reason naming the dominant driver
• Compact, mobile-friendly info label (text, not a table; sized Small/Normal/Large): score, state, direction, break level, distance-in-ATR, reason and a mini-bar per sensor (BOS shown separately as a confirmed event, not mixed into the score)
• Debug log on state change · risk-band colours encode risk, never direction
• Break Pressure / Critical / Top-Break / Bottom-Break alerts

Sensors

Sensor	Role	Reads (uptrend → top-break)
Near Break Level	Location → Trigger	Distance from close to the last swing low, in ATR — risk rises on approach (full within Near Break Level ATR Distance)
Confirmed BOS	Trigger (event, not weighted)	Close below the last swing low — graded by depth in ATR (confirmed break floored at 0.65); shown in the info label, not part of the score
Failed Breakout / SFP	Trigger + Location	High sweeps above the last swing high but the bar closes back below it
Structure Erosion	Structure	The most recent swing high is lower than the prior one — graded by the drop in ATR
Pivot Divergence	Momentum	The latest price-pivot high is higher than the prior, but RSI sampled at those pivots is lower

Each sensor mirrors symmetrically for a downtrend (higher low, break above swing high, bullish divergence, etc.). Structure Erosion and Pivot Divergence fire on the bar the new swing pivot is confirmed; all sensors then decay over Evidence Memory Bars, so the gauge reflects recent structural evidence and old evidence fades instead of being held as long as the last two pivots stay unchanged.

Score and state

risk = Σ(sensorᵢ · weightᵢ) / (Σweight − min(w_Erosion, w_Divergence)) · 100 over the four weighted sensors (capped at 100), EMA-smoothed, gated to zero when no trend context exists. Interpretation: 0–35 Quiet, no acute break risk · 35–60 Watch · 60–80 Break Pressure, structure turning critical · 80–100 Critical — break very near, or Structure Broken once a close confirms beyond the swing with the EMA context flipped (a swing break that came first as Break Pressure is confirmed on the later bar the EMA context flips, if price still holds beyond that swing). Erosion and Divergence exclude each other at the same pivot pair, so the divisor is the reachable maximum and 100 stays reachable. The risk-band colour (neutral → orange → red → dark red) tells you the severity, not the trade direction; which structure is tested is named in words (“Uptrend at risk” / “Downtrend at risk”).

How to use

• Trend management: rising SBR inside a position warns the trend structure is deteriorating before the move fully reverses.
• With TPS: read them together — SBR high + TPS low = M-top / structure break plausible (unclean and breaking); SBR low + TPS low = just range / chop, no clean break; SBR high + TPS high = strong trend running but approaching a critical level.
• Confirmation, not prediction: the Break of Structure event only fires on an actual close beyond the swing — SBR escalates as evidence accumulates rather than calling a top in advance.

Notes

• Symmetric and situation-driven — it never assumes long or short; the active trend decides which structure is tested.
• Default lengths suit swing timeframes (4h/1D). Shorten Swing Pivot Length and Range Window for intraday.