METHODOLOGY

No black box. Every caution has a reason.

NOP Proof Score summarizes observable risk from 0 to 100. The investor brief separately reports research posture, critical conditions and evidence coverage—never a buy or sell instruction.

Contract & exit controls

32%

EVM/Tron honeypot and privileged controls or Solana mint, freeze, close, metadata, fee and hook authorities.

Liquidity & exitability

24%

Aggregated indexed liquidity, top-pool concentration, pool maturity and liquidity relative to FDV.

Wallet concentration

18%

Largest holder, top-10 share, owner/creator exposure and holder breadth before manual entity labeling.

Market behavior

18%

Turnover, buy/sell flow, 24-hour price behavior and cross-pool price dispersion as clues—not manipulation verdicts.

Project evidence

Public website, social and repository claims linked to third-party metadata; claims are not identity verification.

Facts

Values directly returned by a named data source, with a link to inspect that source.

Inferences

Explicit rules applied to facts—for example, shallow liquidity relative to FDV or a young primary pool.

Unknowns

Missing data lowers coverage and can cap the score. It is never silently treated as safe.

Market Intelligence Radar

Radar keeps three evidence layers separate. Narrative Signal Strength measures category market-cap change, reported volume and market depth. Network Hype Velocity combines live 1-hour DEX attention with NOP Fast Lane new-pool acceleration, verified candidate breadth and the observed critical/high-risk share where contract coverage exists. Token candidates are split into Verified Opportunities and Raw Discovery: unverified attention scores are capped at 69/100 and cannot receive a high-confidence label until token-risk enrichment completes. Extreme price expansion is penalized and marked overheated. These measurements prioritize research—they do not predict returns, prove a catalyst or represent every contract on every network.

Rapid Intelligence Pipeline

Fast Lane checks the supported global GeckoTerminal new-pool feed every 15 seconds with exponential provider backoff, stores an immediate market preflight and enriches a bounded batch of risk-research candidates. Empty queues are not polled. Deeper multi-source analysis runs when a researcher searches or monitors a token. This covers newly surfaced tradable pools—not every contract deployed on every blockchain.

Watchlist Alert Intelligence · controlled beta

The rules, baseline comparison, in-app archive and email templates support 6, 12 or 24-hour schedules. The provider-safe launch profile currently keeps the recurring monitoring worker disabled unless an account is admitted to the pilot. When active, NOP Intel separates risk deterioration from strengthening research conditions; these are not buy signals, price forecasts or guarantees of future performance.

Supply Trail

For premium EVM research, NOP Intel resolves a compatible Blockscout explorer and reconstructs indexed token transfers. A dedicated zero-address ledger calculates mint and burn totals; the broader ledger surfaces first recipients, creator/owner gross outflows and the largest movements. Coverage is marked complete only when the index and returned history agree. Transfers use current value for context and are never described as realized proceeds. USD sale proceeds require a matching observed primary-pool swap.

Wallet Flow Forensics

For premium research, recent trades from the primary surfaced pool are grouped by initiating address. A “no sampled buys” label means that the address sold but did not buy the token inside that capped pool window; inventory may still have arrived through transfers, another pool, another chain or earlier activity. “Two-sided / MM-like” requires repeated buys and sells, balanced volume and direction alternation. It is a behavioral similarity—not proof that the address is a market maker, insider or related party—and it does not change the Proof Score.

Decision rules and limitations

A direct transfer/exit constraint caps the score at 18; other hard-stop conditions cap it at 34. Raw holder concentration can include exchanges, bridges, burn or LP addresses and therefore requires manual labeling. Trade imbalance and turnover are screening clues, not proof of market manipulation. Automated analysis can still miss upgrades, related-wallet clusters, compromised keys, cross-chain exposure and new exploit classes.