How to tell if a token is a honeypot
Five checks, in order of how often they catch a bad token. The scanner above runs all five in a few seconds; here is what each one means if you want to do it by hand.
1. Simulate a sell, not just a buy
A honeypot always lets you buy. The only reliable test is to sell. Tools that read the contract's declared "tax" variables miss custom logic; a simulated sell catches it because it measures what actually arrives.
2. Measure the real taxes
Compare what the router quotes with what the wallet receives. Anything above about 10% each way is expensive; above 30% is a trap.
3. Read the owner's powers
If ownership is not renounced, check what the owner can call: setFee, blacklist, pause, setTradingEnabled, mint. Any of these can turn a sellable token into a honeypot after you buy. Upgradeable proxies can change everything.
4. Check whether liquidity can be pulled
Find the main pool, then check who holds the LP tokens. Burned (sent to the dead address) or locked in a known locker is good; held by the deployer means they can remove it.
5. Look at who holds the supply
A deployer wallet with 20% or more of the supply can crash the price on its own.
For bots and agents, the same checks are available as a JSON API and as an MCP tool.
More
BSC honeypot checker · Base honeypot checker · Rug pull checker · How to spot a honeypot · New launches