Solana Token Safety Checks Before You Buy: Mint, Freeze & LP

Correction (September 2, 2026): The 8-of-53 finding is now identified as a descriptive share from a fixed, strategy-selected paper cohort, not a universal rug probability or funded result.

Why do these four Solana safety checks exist?

Every Solana token is an account governed by the SPL Token program (or its successor, Token-2022). The program gives token creators specific powers — and each power, left active, is a specific way to take your money. These checks aren't superstition; they map onto rug mechanisms the screening workflow is designed to reject.

What is mint authority?

Mint authority is the on-chain right to create new tokens of a given mint. Per the SPL Token documentation, whoever holds it can increase supply at will. For a meme coin, an active mint authority means the deployer can print billions of new tokens and sell them into the pool — instant dilution, mathematically guaranteed.

What to verify: mint authority shows as revoked/null on Solscan or in any scanner. In our telemetry this was the third most common rejection reason — 179 tokens flagged in a single week. Legitimate meme coins revoke it at or shortly after launch; there is no innocent reason for a "community token" to keep it.

What is freeze authority?

Freeze authority lets its holder freeze any token account for that mint — after which the frozen wallet cannot transfer or sell. As a rug mechanism it's brutal in its simplicity: let people buy, freeze them, sell into the liquidity they can no longer touch.

What to verify: freeze authority revoked/null. Rare legitimate uses exist (regulated assets, sanctioned-wallet compliance for stablecoins), but for a meme coin an active freeze authority earned 45 rejections in our one-week telemetry and deserves every one of them.

What does "LP burned" mean?

When liquidity is added to an AMM pool, the provider receives LP tokens — a claim on the pool's contents. If the deployer keeps those LP tokens, they can withdraw both sides of the pool at any moment: the classic liquidity rug, and the reason "LP burned" became a meme-trading mantra.

Burning the LP tokens (sending them to an unspendable address) makes the liquidity permanent. Locking them in a time-lock contract is weaker but acceptable if the lock is long and verifiable. Our engine rejected 107 tokens in a week for pools with 0% LP burned or locked.

Pump.fun caveat: tokens that graduate from pump.fun get their LP burned automatically at graduation — that vector is closed. But supply-based rugs (concentrated insiders dumping) survive graduation just fine, which is why holder concentration stays the top flag even on graduates.

What are Token-2022 extensions and transfer hooks?

Token-2022 is Solana's extended token standard. Most extensions are legitimate (metadata, transfer fees for specific designs), but two are honeypot enablers:

• Transfer hooks — developer-controlled code that runs on every transfer and can make sells fail. See Solana's own transfer hook guide for how much power the hook program has. You can buy; your sell reverts. 58 rejections in our telemetry.
• Permanent delegate — an address that can move (or burn) tokens out of any holder's account, forever. For a meme coin this is confiscation-as-a-feature.

What to verify: the token uses classic SPL Token, or if Token-2022, has no transfer hook / permanent delegate / confiscation-grade extensions. Our real-money desk hard-rejects these before any other check runs.

The 4-line Solana token safety checklist

1. Mint authority: revoked.
2. Freeze authority: revoked.
3. LP tokens: burned (or verifiably long-locked).
4. Token-2022 extensions: none hostile (no transfer hook, no permanent delegate).

All four are visible on-chain in under a minute via Solscan, or in one shot in MemeAssist's analyzer, which also layers on holder-concentration analysis and creator-wallet history. New accounts receive 1 free full token analysis with Trade Setup. Further analyses and other paid tools require account credit. In the fixed, strategy-selected Jul 25–Aug 2 paper cohort, 8 of 53 modeled positions were rug-tagged after the full automated gate; that descriptive share is not a population rate or a funded result. See the cohort timing and limitations.