CIF-Score-v1.0
Methodology v1.0.0
The scoring contract behind CIF. Every future production snapshot must be reproducible against one exact methodology version.
● DRAFT-LOCKED
Core rules
No black-box scoring.
• Quality, timing, risk and confidence are independent scores.
• A score is never a probability unless separately calibrated on historical outcomes.
• Every scored criterion must have evidence, a timestamp and a source class.
• Missing applicable data lowers confidence; it must not silently receive a neutral score.
• N/A criteria are allowed only when structurally inapplicable and their weight is renormalized within that score.
• Methodology version is stored with every future snapshot so historical results remain reproducible.
AQS
Asset Quality Score
sum(component_score_0_100 × component_weight)
Product & utilityReal problem, product usefulness, token necessity, product maturity.
15%Adoption & tractionUsers, transactions, fees/revenue/TVL where applicable, retention and growth quality.
15%Development & executionRelease cadence, repository activity, contributor breadth, roadmap delivery.
10%Competitive positionDifferentiation, switching costs, integrations, market share and defensibility.
10%TokenomicsSupply schedule, FDV/float, utility, value capture, emissions and distribution.
20%Liquidity & market accessReal spot liquidity, depth, venues, volume quality and slippage.
10%Security & resilienceAudits, exploit history, admin controls, bug bounty and operational resilience.
10%Governance & decentralizationControl concentration, upgrade powers, validator/governance distribution and key-person dependence.
10%MOS
Market Opportunity Score
sum(component_score_0_100 × component_weight)
1M / 1W structureLong-term structure, regime, major support/resistance and higher-timeframe trend.
15%1D trendDaily market structure, EMA regime, breakout/reclaim quality and trend persistence.
15%4H / 1H structureEntry-window structure, HH/HL or LH/LL, BOS/CHoCH and local invalidation.
15%Volume & participationBreakout participation, volume expansion/contraction, absorption and follow-through.
15%Relative strengthPerformance versus BTC, ETH and relevant sector benchmark over multiple horizons.
10%Valuation / range positioningDistance from major supply, prior range, ATH and risk/reward asymmetry.
10%Derivatives positioningFunding, open interest, basis and crowded positioning where derivatives exist.
10%Near-term catalystsVerified events with a defined date/window and plausible market relevance.
10%RS
Risk Score
sum(component_risk_0_100 × component_weight), then apply critical flags
Liquidity risk
15%Unlock & inflation risk
15%Holder / control concentration
15%Smart-contract / security risk
15%Protocol & infrastructure dependency
10%Governance & key-person risk
10%Regulatory / legal risk
10%Volatility & market-structure risk
10%Critical risk flags• Unresolved critical exploit or credible insolvency concern• Unverified or mutable contract risk that can materially alter balances/supply• Single actor can freeze/mint/drain without meaningful safeguards• Known near-term supply event large enough to invalidate normal scoring assumptions
CS
Confidence Score
0.40 × coverage + 0.30 × freshness + 0.30 × source_quality
Evidence coverageShare of applicable score weight backed by evidence.
40%Data freshnessAge versus expected refresh interval for that evidence type.
30%Source qualityPrimary, first-party, high-quality aggregator or secondary source class.
30%Evidence quality
Source classes
A
100/100Protocol docs/contracts, exchange API, chain data, audited filings, official repository.
B
85/100Established data provider with transparent methodology/API.
C
70/100Reputable secondary research with traceable underlying sources.
D
50/100Unverified secondary claim, social post or manually asserted value.
Validation
Historical calibration
Snapshots are evaluated after 30, 90, 180, 365 days. Until that calibration exists, CIF scores are ordinal decision scores — not probabilities.
Measure whether score bands have predictive/decision value and recalibrate weights only from out-of-sample evidence.