Honeypot checkers do their job well. Paste a token, learn whether the sell path reverts — and
the good ones, including the free public ones, get that right. The problem is that the honeypot
question is one of five, and the other four are where most money is actually lost.
§01 · the honest part
The free tools are good. Use them.
If your question is literally “can I sell this token”, the established scanners answer it, and several answer it for free — Honeypot.is exposes a public API with no key, and GoPlus returns a broad field set at no cost. There is no point pretending otherwise, and a page that told you they were inaccurate would be lying about something you can verify in thirty seconds.
What follows isn’t an argument that they get the honeypot question wrong. It’s an argument that the honeypot question is too small.
§02 · exit size
A simulation sells dust. You won’t.
Every buy/sell simulation round-trips a small probe — it has to, or the probe itself moves the market. So sellable: true means a tiny amount came back out. If the pool holds four thousand dollars and you arrive with one, you will lose a large fraction to slippage on the way out. No trap, no malicious code, same result.
That’s why depth belongs next to the verdict, in dollars, not as a boolean. A partial trap is worse still: a max-transaction limit lets the small probe through and blocks real size, so a single-size check reports “sellable” on a token nobody can exit.
§03 · supply
Mint capability dilutes you legally.
A token with an open mint function doesn’t need to block your sell. The deployer prints supply, the price collapses, and every transaction along the way is perfectly valid. The sell path stayed open the whole time — the honeypot check was right and you still lost.
§04 · the human
Who shipped it, and what did they ship before?
Contracts don’t rug; deployers do. A wallet created an hour ago that has already deployed nine tokens is a different proposition from one with two years of history, and no amount of bytecode analysis will tell you which you’re dealing with. Deployer age and prior-token history are the cheapest strong signal available, and almost no checker surfaces them.
§05 · right now
The check is a snapshot. The rug is an event.
Every scan describes the token at the instant you asked. Liquidity removal happens later, which is precisely why it works. A check that also reports recent liquidity-removal events and whale exits on this token is telling you something a snapshot cannot: that the exit is happening while you read.
§06 · all five, one call
What that looks like.
true402 evaluates the five together on Base and returns a single verdict — simulation, liquidity in USD, mint and ownership, deployer reputation, and live rug/whale activity. Free to try in the browser with no wallet at the Base token scanner, and available as a CLI, an HTTP API, or an MCP tool inside your agent.
one call, five signals
# One call, all five signals — Base. ~$0.01 USDC over x402, no API key.
curl -s -X POST https://true402.dev/api/v1/base/token-report \
-H 'content-type: application/json' \
-d '{"token":"0x<token-address>"}'
# 200 →
"verdict": { "rating": "caution", "score": 70, "reasons": ["mint_function_present"] },
"safety": { "honeypot": { "simulated": true, "sellable": true, "roundTripBps": 9801 },
"liquidity": { "usd": 14144843 }, // can you actually exit at size?
"mintable": true, "ownership": "none" },
"recentPulls": [], // liquidity being removed right now
"recentWhaleSwaps": []
§07 · faq
Questions.
Is Token Sniffer accurate?
For what it measures, broadly yes — and the same is true of Honeypot.is, GoPlus and the other established scanners. They detect the classic trap (buy works, sell reverts) and report buy/sell tax reliably on the chains they cover well. The accuracy question is usually the wrong one: a checker can be perfectly accurate about the honeypot question and still leave you exposed, because most money is lost to tokens that pass a honeypot check at the moment you look and are drained afterwards.
What does a honeypot check not tell you?
Four things, all of which have taken more money than classic honeypots: whether liquidity is deep enough to exit at your size, whether the contract can mint new supply and dilute you, who deployed it and what they shipped before, and whether liquidity is being pulled right now. A token can be perfectly sellable in a simulation and still be a rug ten minutes later when the deployer removes the pool.
What is the best honeypot checker?
For a one-off "can I sell this", the free public tools are good and you should use them — Honeypot.is exposes a free keyless API and GoPlus returns a broad field set at no cost. The case for anything else starts when you need more than the honeypot answer in one call, need it inside an agent or bot rather than a browser, or need liquidity, mint capability, deployer history and live rug activity evaluated together.
Why does liquidity depth matter more than the honeypot flag?
Because sellable is not the same as exitable. A simulation typically round-trips a tiny probe amount. If the pool only holds a few thousand dollars, that probe succeeds while your real position moves the price so far that you lose most of it on the way out — no honeypot required. Any check that reports sellable without reporting depth in dollars has answered half the question.
How do I check a Base token for all of these at once?
true402 returns them in one call: the buy/sell simulation, liquidity depth across Uniswap V3 and Aerodrome pools in USD, ownership and mint capability, deployer reputation, and recent liquidity-removal and whale activity on that token — as a single avoid/caution/ok verdict. It is free to try on the web with no wallet, and available as a CLI, HTTP API and MCP tool for agents.