Check a token
Methodology · Token risk

How HostDeFi rates token risk

Every risk score on HostDeFi comes from the same automated screen: a set of on-chain and market signals read at the moment you check, weighed together into a rating and a list of specific flags. This page explains what those signals are — and, just as importantly, what a score can't tell you.

The signals

Contract authoritiesWhether the token's mint authority and freeze authority are revoked or still live. A live mint authority means supply can be inflated at will; a live freeze authority means your tokens can be locked. Revoked authorities are the baseline of a token that can't be rugged by its own contract.
Liquidity depth and statusHow much real liquidity backs the token, how it's split across pools, and whether pool tokens are burned or locked. Thin or unlocked liquidity is the single most common exit route in a rug pull.
Holder concentrationWhat share of supply sits with the top holders and the deployer, and whether early buys cluster in ways typical of sniper rings or insider allocations. Concentrated supply means a few wallets can move the price at will.
Deployer and team-wallet behaviorThe deploying wallet's history: prior token launches, funding trails, and whether team-linked wallets have been distributing into buyers.
Trading-pattern checksSell-side simulation to catch honeypots (tokens you can buy but not sell), abnormal buy/sell taxes, and volume patterns consistent with wash trading rather than organic interest.
Metadata and ageToken age, naming that impersonates established projects, and whether the basic metadata checks out. Very new tokens with borrowed names score cautious by default.

Data sources

Signals are read from public on-chain data on the token's own network, combined with live market data (prices, liquidity, and volume) from public market-data providers. Scores reflect the state of the token at the time of the check — liquidity can be pulled and authorities can change after you look, which is why the timestamp matters.

What a score can't tell you

A risk score is a screen, not a guarantee. It catches the mechanical patterns of past scams; it cannot prove a team's intentions, and new scam designs appear constantly. A clean score means no known red flags were visible at check time — not that a token is safe.

HostDeFi's ratings are informational tooling, not an audit and not financial advice. Never trade more than you can afford to lose, and treat every new token as risky regardless of its score.

Programmatic access (API)

The exact scoring described here is available over HTTP through the HostDeFi Token Risk API — the API returns the same A+–F grade, the same knowns/unknowns receipts, and the same refusal to grade a token that is too fresh or too under-described to judge honestly (the response's graded field is false in that case, with a stated reason). On Solana the API additionally returns the pre-graduation on-chain signals — bundled-launch detection, curve-aware holder concentration and Token-2022 trap flags — read directly from the mint. The API grade and the on-page grade are produced by the same engine, so they never disagree, and the flags' measured hit-rate is published on the live backtest.

Versioning & authorship

This methodology is maintained by HostDeFi Research and versioned in the open — the scoring engine's rules live in the public repositories, and every research report names the exact data sources and snapshot date it was computed from. Current methodology version: v2 (August 2026) — added verified-token requirements for contract-address claims, liquidity/holder floors for token resolution, and quality gates before any generated page publishes. Corrections: @hostdefiapps.