> ## Documentation Index > Fetch the complete documentation index at: https://docs.dequan.xyz/llms.txt > Use this file to discover all available pages before exploring further. # Liquidity Drain Detection The textbook Solana rug works like this: the deployer (or an insider) pulls liquidity out of the token's pool faster than the market can react. By the time the chart shows a price collapse, the pool is already empty and exits have stopped working. The defence against this is simple and absolute: **watch the pool's actual on-chain SOL reserves**, not just the price feed. That is what Dequan does. ## Real reserves, not derived signals Most tools infer liquidity from price action — a sudden drawdown is treated as a possible drain. The problem with that approach is that **price moves before the inference fires**, which means you're alerted *after* you've already taken the loss. Dequan reads the underlying pool's SOL reserves directly from chain. When those reserves move materially down — and especially when they drop in a single transaction by an amount consistent with a deliberate drain — the alert fires within seconds, often before the price feed has fully repriced. That window — between the on-chain drain and the price catching up — is the window where you can still exit. ## Hysteresis, not single-reading panic A naive drain detector would fire on every transient dip and drive everyone insane. Dequan's detector requires **two consecutive readings** below threshold before raising an alert. This is deliberate: it eliminates false alarms from a single noisy reading or a brief liquidity migration, while still firing fast on actual drain events (which produce sustained low readings, not bouncing ones). The result is an alert system that **stays quiet until something is really wrong, then becomes very loud very fast.** ## What you see when it fires A red banner appears across the cinematic chart. The candle palette shifts to warning colors. Existing positions show a "drain detected" annotation. Any open positions in the affected token gain a blinking red badge in the drawer header and on the candle chart. The recommended action becomes immediate exit. The token's glyph turns red and is pushed visually toward the field's "decay" zone. Other users browsing the Pump Zone see the same warning instantly. A notification appears with the token, the magnitude of the drain, and a one-tap exit button. ## Auto-recovery If the reserves come back — which happens sometimes during legitimate liquidity migrations or temporary market-maker activity — the detector clears automatically once readings are healthy again. The alert is **not** sticky beyond what conditions warrant. This means the detector can be trusted to be *current* — a red banner that's still up means the pool is *still* in trouble, not that it once was an hour ago. ## What it doesn't catch Drain detection is one specific defense against one specific attack pattern. It does not protect against: * **Slow rugs** where liquidity is removed gradually over hours * **Sandwich attacks** where you take MEV losses on an otherwise healthy pool * **Honeypots** where you can never sell in the first place — handled by [Honeypot Guard](/safety/honeypot-guard) * **Insider dumps** where the dev sells a large supply allocation onto your bid — visible on the [Top Traders](/charts/token-detail-tabs) tab Different attacks, different defences. Drain detection is the dedicated defence against the most common one. ## Why this matters In our internal testing across many months of meme-token launches, classic drain rugs are the single most common cause of catastrophic losses for traders who otherwise picked decent tokens. The losses come from being asleep at the wheel — being on a chart for *one* moment too long after the rug fires. Dequan's detector is designed to fire **before you've finished forming the thought "wait, what's happening to the price?"** That is the entire point. Live monitoring of concentration, snipers, and dev wallet behaviour.