Every threshold and weight the scoring uses, live from atlas_foundation.atlas_thresholds. Edit within each knob’s allowed range, Save, then Preview the impact and Commit to re-blend the live composite from the cached lens scores — seconds, because only the weighting re-runs. 153 active knobs.
Weight of the Catalyst lens in composite scoring
Weight of the Flow lens in composite scoring
Weight of the Fundamental lens in composite scoring
Weight of the Policy lens in composite scoring
Weight of the Technical lens in composite scoring
Weight of the Valuation lens — neutral descriptor, acts as multiplier not additive weight
Convergence multiplier when exactly 2 lenses are firing
Convergence multiplier when exactly 3 lenses are firing
Convergence multiplier when 4 or more lenses are firing
Minimum rescaled lens score to count as firing for convergence bonus
Minimum number of firing lenses required for High conviction tier
Minimum composite score for High conviction tier
Minimum number of firing lenses required for Highest conviction tier
Minimum composite score for Highest conviction tier
Minimum composite score for Medium conviction tier
Minimum composite score for Watch conviction tier
Composite multiplier applied in Cheap zone
Valuation score threshold above which zone is Cheap
Composite multiplier applied in Deep Value zone
Valuation score threshold above which zone is Deep Value
Composite multiplier applied in Expensive zone
Valuation score threshold above which zone is Expensive (below = Overvalued)
Composite multiplier applied in Fair zone (neutral)
Valuation score threshold above which zone is Fair
Composite multiplier applied in Overvalued zone (below Expensive threshold)
Bottom-quintile cutoff for RS percentile rank within tier
Top-quintile cutoff for RS percentile rank within tier
Maximum |30-week MA slope σ| for flat MA condition in Stage-1 base
Minimum weak-state weeks (out of last 10) for Stage-1 base qualification
Maximum participation_RS for Avoid sector
Minimum participation_RS for Overweight sector
Bottom cross-sector RS percentile cutoff for sector_state = Avoid
Top cross-sector RS percentile cutoff for sector_state = Overweight
Maximum participation_RS for Underweight sector
Lookback window for ema_10 at high/low detection
Maximum |ema_10_ratio - ema_20_ratio| for Flat (EMAs converged)
Maximum |ema_10_ratio - 1| for Flat classification
Minimum effort_ratio_63 for Accumulation
Minimum volume_expansion for Accumulation
Maximum effort_ratio_63 for Distribution
Maximum effort_ratio_63 for Heavy Distribution
Accumulation: moderate up/down delivery asym (pp)
Accumulation: strong up/down delivery asym (pp)
Accumulation: high sustained delivery (avg30)
Accumulation: low sustained delivery (avg30)
Accumulation: moderate 30d-vs-60d trend
Accumulation: strong 30d-vs-60d trend
Flow weight: delivery accumulation
Flow weight: promoter
Flow weight: smart-money
Days after a fiscal year's period_end before the annual balance sheet is treated as knowable (PIT as-of proxy).
Days after a quarter's period_end before its income statement is treated as knowable (PIT as-of proxy).
Minimum extension % for High risk
Maximum extension % for Low/Normal risk
Minimum vol_ratio_63 for High risk
Maximum vol_ratio_63 for Low risk
Maximum vol_ratio_63 for Normal risk
Minimum trading days of OHLCV history before classification eligibility
Minimum trailing 60-day median daily traded value for liquidity gate
Minimum 30-week MA slope (σ-normalized) for flat-or-rising condition
Multiple of 252-day median vol above which dislocation activates
Maximum India VIX for Cautious regime
Minimum pct_above_ema_50 for Constructive regime
Maximum India VIX for Constructive regime
Width of 'near EMA 200' band for Cautious regime trigger
Maximum pct_above_ema_50 for Risk-Off regime
Minimum pct_above_ema_50 for Risk-On regime
Maximum India VIX for Risk-On regime
Breadth (% of stocks above 200dma) below which regime moves to Below-Trend (placeholder).
Breadth (% of stocks above 200dma) below which regime moves to Risk-Off (placeholder).
Cross-sectional dispersion above which regime moves to Elevated (placeholder).
Smallcap RS z-score below which regime moves to Below-Trend (placeholder; Phase 0.5h-prime sweep replaces this).
Smallcap RS z-score below which regime moves to Risk-Off (placeholder; Phase 0.5h-prime sweep replaces this).
VIX percentile above which regime moves to Elevated (placeholder).
VIX percentile above which regime moves to Risk-Off (placeholder).
ETF rs_3m_benchmark cutoff (in pct) above which etf rs_state becomes Strong
Top-percentile cutoff (per category) for ETF is_atlas_leader=TRUE
Maximum AUM (Cr) for full ETF AUM score
Minimum AUM (Cr) for full ETF AUM score
ETF composite — AUM bracket weight
ETF composite — expense ratio weight (inverse)
ETF composite — liquidity weight
ETF composite — matrix conviction weight
ETF composite — sector strength weight
ETF composite — tracking quality weight
Minimum AUM in Overweight/Neutral sectors for Aligned
Maximum AUM in Avoid sectors for Aligned
Minimum AUM in Leader/Strong/Emerging stocks for Strong-Holdings
Maximum AUM in Weak/Laggard stocks for Decent classification
signal_score < this → Poor decision state
signal_score >= this → Sharp decision state
Min |weight_delta| (%) to classify as increase/decrease vs noise
Top-percentile cutoff (per category) for is_atlas_leader=TRUE
Maximum AUM (Cr) for full cost+manager AUM score (above = liquidity drag)
Minimum AUM (Cr) for full cost+manager AUM score
Bottom-percentile cutoff (per category) for is_avoid=TRUE
Top-N holdings used for the holdings conviction aggregation
Years of NAV history required before confidence_low=FALSE
MF composite — cost + manager layer (TER/tenure/AUM/age)
MF composite — holdings conviction layer (24-cell weighted avg)
MF composite — risk-adjusted return layer (Sharpe/Sortino/Alpha/MaxDD/Calmar)
MF composite — style + sector tilt layer
Min entry+exit count to compute signal_score
Maximum distance from 20-EMA (%) for breakout entry trigger
Contraction narrowing lookback bars
Contraction: lookback bars for highest high resistance
Contraction: max % from highest high
NPC: close must be in bottom X% of range
NPC: TRP ratio threshold
NPC: max RS cross-sector percentile
NPC: min stage to fire (3=topping, 4=declining)
NPC: volume vs 20-bar avg
PPC: close must be in top X% of range
PPC: max pct below 52-bar high to qualify
Morales: lookback bars for down-day volume max
PPC: TRP ratio threshold
PPC: min RS cross-sector percentile
PPC: min Weinstein stage (2=advancing)
PPC: volume vs 20-bar avg
Stage 1B: max % below SMA_150 to qualify
SMA slope lookback days
Weinstein SMA period (trading days)
Min TRP% for tradeable stock
Hard cap on desk orders booked per nightly cycle
Hard cap on desk holdings per sector
Top-N by composite fed to the desk agents
ETF BUY cost: stamp 0.015% + txn/GST
ETF SELL cost: STT 0.001% + txn/GST
MF BUY cost: stamp duty 0.005%
Equity MF redemption STT 0.001%
Equity delivery BUY cost: STT 0.1% + stamp 0.015% + txn/SEBI/GST
Equity delivery SELL cost: STT 0.1% + txn/SEBI/GST
Default initial capital for a new portfolio
Minimum days between policy changes (anti noise-chasing)
Challenger must beat champion by this many pp of excess return on validation
Minimum train-window trades for a candidate to be considered
Walk-forward TRAIN window for system-portfolio policy search
Out-of-sample VALIDATION window (never trained on)
Holding days below which the fund exit load applies
Equity-MF exit load on redemptions within the load window
Default max position size as a fraction of portfolio value (slots = floor(1/pct))
Holding days threshold for LTCG on listed equity/MF
LTCG exemption per financial year (per portfolio)
LTCG rate on equity/equity-MF above the FY exemption
STCG rate on equity/equity-MF (held < LTCG threshold)
Trend pts when EMA21>EMA50>EMA200 (full bullish stack)
Trend pts for a partial EMA alignment (EMA21>EMA50 or EMA50>EMA200)
Price-vs-EMA200 fraction above which trend reads 'strong above'
Price-vs-EMA200 fraction below which trend reads 'well below'
RS pts when EMA21>EMA50 (medium-term up)
RS pts when EMA50>EMA200 (golden cross / long-term up)
1-week return above which EMA21 slope reads 'steep up'
1-week return below which EMA21 slope reads 'steep down'