Technical names stay visible for auditability. These cards explain what each module actually does. Confidence means checks passed; it is not a probability of profit.
Core Scorecard
CORE-SCORECARD/v2.1.0
- Name
- Directional checklist
- Input
- OHLCV, EMA, RSI, ATR, volume and price structure.
- Purpose
- Count aligned conditions consistently.
- Output
- LONG, SHORT or WAIT plus passed and failed checks.
Limit: the score is not calibrated odds and can lag sudden regime changes.
SARVAM-369
SARVAM-369/v1.0.0
- Name
- Final Decision Engine
- Input
- Validated outputs from the scorecard and specialist engines.
- Purpose
- Combine evidence and veto weak or conflicting direction.
- Output
- Final educational LONG, SHORT or WAIT label with reasons.
Limit: thresholds are reasoned, not proven out-of-sample; it may only demote a direction to WAIT.
CWIF-369
Included in TESBOT-STACK/v3.5.0
- Name
- Setup Quality Filter
- Input
- Trend, volatility, candle structure and proposed paper levels.
- Purpose
- Reject noisy, late or internally inconsistent setups.
- Output
- Pass, veto-to-WAIT and named reasons.
Limit: filtering can miss moves and does not guarantee higher win rates.
TRAMSE-369
TRAMSE-369/v2.0.0
- Name
- Regime and Volatility Context
- Input
- Historical candle returns and ranges.
- Purpose
- Describe trend, transition and volatility conditions.
- Output
- Regime label, filtered trend and volatility estimate.
Limit: fixed parameters were not fitted as a production forecasting model and can lag shocks.
MARP-RX
MARP-RX-369/v1.3.0
- Name
- Risk, Liquidity and Cost Layer
- Input
- Returns, tail observations, public order book and declared costs.
- Purpose
- Test whether a paper edge survives risk and cost checks.
- Output
- Tail-risk estimate, liquidity grade and take/reduce/skip context.
Limit: public order books are not an executable consolidated venue and historical tails can understate crashes.
RAUCL-MI
RAUCL-MI-369/v1.2.0
- Name
- Uncertainty Layer
- Input
- Residuals, checklist alignment and regime context.
- Purpose
- Make uncertainty and data reliability explicit.
- Output
- Adjusted research score, uncertainty band and drift warning.
Limit: the adjusted score is a heuristic, not a fitted probability calibration.
ADV-369
ADV-369/v1.1.0
- Name
- Advanced Research Approximation
- Input
- Price, volatility and market-structure features.
- Purpose
- Provide educational options, rough-volatility and microstructure context.
- Output
- Research descriptors and WAIT-aware action label.
Limit: simplified approximations only; not a trained reinforcement-learning policy or executable pricing system.
CAPTA-369
CAPTA-369
- Name
- Historical Pattern Analysis
- Input
- Timestamped historical exchange candles.
- Purpose
- Compare current structure with past formations without look-ahead.
- Output
- Pattern dates, comparisons and observed historical outcomes.
Limit: historical resemblance does not imply recurrence or future profit.
KNIE-369
KNIE-369
- Name
- News Impact Context
- Input
- Available public and official news feeds.
- Purpose
- Attach source-labelled event context to research.
- Output
- Timestamped items, regions and impact descriptors.
Limit: feed coverage and latency vary; missing news is never treated as proof that no event exists.
NCIR-369
NCIR-369/v1
- Name
- New Coin Intelligence Radar
- Input
- Available listing, contract and market evidence.
- Purpose
- Organize early-asset research and flag missing evidence.
- Output
- Research status, evidence list and NO SIGNAL where safety checks fail.
Limit: early assets have extreme data, liquidity and fraud risk; the radar never promises a listing or price move.
