Each line is one live pool's liquidity, as % change over the last 90 seconds (0% = where it stood 90s ago). Green climbs as liquidity builds; red dives as it drains. A line falling toward −100% means a pool is being emptied: a rug, as it happens.
Fastest-rising liquidity
…
Pools whose real reserves are climbing fastest right now, measured live on the reserve stream.
Rug watch
…
Small, recently-created pools building liquidity unusually fast: the classic pre-rug profile. High-risk, not a guarantee, not financial advice.
Just drained
…
The experiment, live
Hypothesis: pools whose liquidity rises fastest are the ones that drain. We log every pool that enters the rug-watch profile, then count how many actually drain.
gathering data…
Built on free, public DexPaprika data
No private endpoints, no special access, no API key. Everything you see here runs on the same public data any developer can call today, from two free, keyless DexPaprika endpoints.
POST streaming.dexpaprika.com/sse/reservesLive reserves over one SSE connection, multiplexing every pool on the chart, so a drain shows up the block it happens. This is what turns into the red and green lines above.GET /networks/{chain}/pools/filterREST pool discovery, so you pick which pools are worth watching before subscribing to their reserves.
That's the whole stack: those two calls plus one Cloudflare Worker and a Durable Object you deploy with a single command. Don't take our word for it. Call the endpoints, diff the numbers, then fork the repo or read the docs. If this is what the free data does, picture what you'd build on top of it.