What goes into the score
Can you sell? (hard gate)
Honeypot / can't-sell / can't-buy sends the score straight to CRITICAL, the trap that matters most.
Liquidity lock
We sum the LP that's locked or burned. Unlocked liquidity a deployer can pull is the #1 rug mechanic.
Concentration
Owner/creator % and the top-10 wallets' share, how easily a few holders can dump on you.
Wash-volume proxy
24h volume ÷ liquidity. When volume dwarfs the pool, the "activity" is often self-dealt. (Proxy: see the note below.)
Owner powers
Mintable supply, editable balances, hidden/ reclaimable ownership, self-destruct, pausable transfers, blacklists.
Depth, tax, age & verification
Pool depth, buy/sell tax, how new the pair is, and whether the source is verified.
On honesty: the wash-volume figure is a ratio proxy, public aggregators don't expose unique-trader counts, so true self/circular-trade detection is a paid Phase-2 upgrade. We label it as a proxy rather than overclaim. Full fake-farming (staking-contract APR/TVL) analysis is also Phase-2.
FAQ
How does the Launch Safety Score work?
Starts at 100, subtracts weighted points per risk found (honeypot is a hard gate). Every deduction shows its point cost so you can audit the number.
How do you detect a rug pull early?
The structural tells: liquidity not locked/burned, a deployer or top wallets holding a big share, mintable supply, and volume that dwarfs liquidity. All flagged before you buy.
Does a high score mean it's safe?
No: it means no major structural red flags in on-chain + market data. A good score lowers risk; it's not a guarantee. DYOR.