How to Rug-Check a Crypto Token: A 9-Step Field Guide

Why every memecoin trade should start with a rug check

Memecoins, AI agent tokens, and most DEX-launched tokens are unaudited. Most are deployed with copy-paste contract templates. A meaningful fraction — even on the major chains — are deliberately malicious. The default position should be: every token is a rug until you've checked the fingerprints.

Rug checking isn't paranoia, it's hygiene. The checks are systematic, they take five to fifteen minutes manually, and they catch 80–90% of the rugs that drain retail wallets every week. The other 10–20% are sophisticated, governance-style rugs (legitimate launches that flip malicious after raising) — those require ongoing monitoring rather than a single check.

The nine checks below are what professional traders run before any DEX position. The Sharpe rug check tool automates all nine against any address on Solana, Ethereum, Base, BSC, Arbitrum, or Polygon. The walkthrough below is the manual version so you understand what the tool is doing — and so you can rug-check tokens on chains the automated tool doesn't yet support.

The nine checks

1. Verify the contract source code is published. Open the chain explorer (Etherscan, Solscan, BaseScan, BscScan, Arbiscan, Polygonscan). Search the token contract address. The contract should have a green checkmark or "Source Code" tab — meaning the deployer published the source for public audit. Unverified contracts could contain anything; treat as yellow flag minimum, red flag for anything you don't have prior conviction on.

2. Search for mint functions. In the verified source, search for the strings mint, _mint, or mintTo. If a public function with mint in its name is callable by the deployer or owner, the supply can be diluted at any time. Stablecoins (USDC, USDT) have controlled mint and that's fine. For a memecoin, mint capability is usually a red flag.

3. Check the deployer wallet share. On the chain explorer, click "Holders" on the token page. Find the deployer address (often labeled or the first holder created at deploy). Check what percentage of total supply they still hold.

4. Verify liquidity is locked. Click on the LP token (the Uniswap V2/V3, PancakeSwap, Raydium pair) on the chain explorer. Check who holds the LP tokens. Three patterns are safe:

• LP tokens are sent to a verified locker (Unicrypt, Team.Finance, PinkSale Lock, or chain-native equivalents) with a documented lock duration of at least 6 months.
• LP tokens are sent to a burn address (0x0000...0dEaD on EVM, or the equivalent system burn address on Solana). Liquidity can never be withdrawn — the strongest possible guarantee.
• LP tokens are held by a multi-sig wallet with documented governance.

If LP tokens are held by an EOA (single private key wallet), the deployer can drain the pool at any moment. Red flag.

5. Test for honeypot. A honeypot is a contract where buys go through but sells are blocked or taxed at 90%+. Manual test:

• Simulate a sell on Tenderly using the deployer's RPC, OR
• Use a free honeypot scanner (Honeypot.is for EVM, RugCheck.xyz for Solana), OR
• Submit through Sharpe's rug check which runs the simulation automatically.

A pure honeypot is unsellable at any price. A "soft honeypot" lets small sells through but blocks larger ones. Both are red flags.

6. Audit the fee structure. Find the contract's _transfer or _takeFee function. Read the buy tax (fee on buys) and sell tax (fee on sells). Three patterns:

• Equal buy/sell, ≤5%: typical for legitimate tokens.
• Equal buy/sell, 5-15%: aggressive but seen in some legitimate tokens (e.g. fee-on-transfer reflection tokens).
• Asymmetric (low buy, high sell): scam pattern. Always red flag.

A token with 1% buy tax and 30% sell tax is buyable but not profitably sellable.

7. Inspect holder concentration. Top 10 holders combined share, excluding burn addresses and the LP contract:

• Below 20%: healthy distribution.
• 20-30%: typical for newer tokens.
• 30-50%: consolidated — single-whale risk.
• Above 50%: dangerous. A few holders can dump and crash the price.

Cross-reference top holders against known team wallets, treasury contracts, and CEX cold wallets. Genuine large holders can be benign (treasury, CEX) but should be transparent.

8. Look for blacklist or freeze functions. Search the source for blacklist, excludeFromTransfer, freeze, or pause functions accessible to the deployer. These let the deployer block specific addresses — including yours after you've bought. Stablecoins like USDC and USDT have blacklist functions for regulatory compliance, which is acceptable. For a memecoin or DEX token, a blacklist function is a yellow-to-red flag.

9. Cross-check against scam databases. Run the contract address through:

• GoPlus (gopluslabs.io) — comprehensive on-chain risk API.
• ScamSniffer (scamsniffer.io) — phishing and scam database.
• CertiK Skynet — security audits and incident database.

Sharpe's rug check runs against several of these simultaneously. A recent match in any database is a definitive red flag — exit even if all other checks pass.

Reading the composite risk score

If you're using Sharpe, the rug check returns a 0-100 risk score combining all nine checks:

The full methodology — how each check contributes to the score, the exact thresholds, the bot-filtering rules — is described in this guide and runs live at /rug-check. Open methodology is a deliberate choice; gating it would prevent traders from sanity-checking the score.

Common rug pull patterns by chain

Different chains attract different rug styles:

Solana memecoins. Pump.fun launches are the highest-volume breeding ground. The most common scam pattern: launch on Pump.fun, graduate to Raydium when market cap hits the threshold, dump the deployer allocation immediately into the new liquidity. The Sharpe Solana rug check flags Pump.fun-graduated tokens and tracks deployer behavior across the launch lifecycle.

Base / Ethereum memecoins. The Clanker and Base ecosystem has attracted both legitimate and rug-prone deployments. The patterns to watch: short-lived liquidity locks (under 30 days), high asymmetric sell taxes, anonymous deployer wallets connected to past rugs.

BSC tokens. PancakeSwap launches still dominate scam volume. The patterns: copy-paste reflection contracts with hidden whale-dump mechanisms, owner-only mint, removable liquidity. Manual rug checks are slower on BSC because the explorer UX is rougher; Sharpe's BSC rug check automates the workflow.

Arbitrum and Polygon. Lower scam volume than other chains — fewer copy-paste exploiters, more legitimate launches. Still run the nine checks; just expect a higher pass rate.

Things automated checks can miss

Manual review still adds value for sophisticated rugs:

Proxy contract upgrades. A contract that's an ownable-upgradeable proxy can have its logic replaced post-launch. The current source might be safe; the next version after upgrade could be malicious. Automated checks flag the proxy presence but can't predict future upgrades. Manually evaluate whether the upgrade governance is decentralized (multi-sig with 6+ signers) or single-key.

Off-chain coordination. If the deployer team coordinates a dump through Discord or Telegram pre-announced, the contract might pass all on-chain checks but the social coordination still rugs holders. Automated tools can't see this. Vigilance on social channels is required for high-conviction holds.

Time-locked malicious logic. Some scam contracts include malicious functions that activate only after a block-number or timestamp. The contract is honest at deployment, scam at unlock. Manual code audits catch this.

Liquidity migration scams. Deployer announces a "v2" migration, asks holders to move funds to a new contract — which is the actual rug. Off-chain awareness of this pattern is the only defense.

For high-stakes positions (>$10K), don't rely solely on automated rug checks. Read the contract, check social channels, and consider your exit liquidity.

Where to go from here

If you're sizing a DEX position right now, run the rug check at /rug-check. Paste the address, get the 0-100 score in 30 seconds. If it scores below 30, you've passed the basic safety bar; size accordingly. If 30-60, do the manual checks above. If above 60, walk away.

For high-frequency DEX trading, every position should pass through the rug check workflow. The cost is 30 seconds; the savings are catastrophic loss avoidance.

The tool is free, the methodology is open, and coverage spans the six major chains. The walkthrough above is the manual version — same checks, more time. Whichever you choose, do them every time.