The pool gets
a say.

Exit Duck is a liquidity drill for two observed meme pools on BNB Smart Chain. Pick a pool, enter a bag size and set a loss tolerance. The drill estimates the quote asset a constant-product pool would return under explicit assumptions.

What the duck measures

The original mark is your amount multiplied by this pool’s reserve ratio. The modeled exit applies your assumed input-token tax, the 0.25% swap fee and integer constant-product arithmetic. It is a single-pool, single-hop model. It does not use a dollar price or search across routes.

For an LP-removal scenario, both reserves shrink by the same percentage while your bag stays the same. The inverse solver finds the largest removal, from 0% to 99%, whose modeled proceeds stay within your loss tolerance relative to the original mark. It resolves to one basis point.

What an observation proves

Each observation checks the network, pool factory, factory registration, token identities and decimals. Reserves and three router quotes are read at the same block, twelve blocks behind the head. The block hash is checked again before accepting the observation. All three integer outputs must match the deployed router.

Observations refresh approximately every five minutes when the worker is available. Shared drills pin a saved observation and assumptions. “Check fresh reserves” keeps those assumptions and requests a newer observation; a cached observation may be reused for up to five minutes.

What it cannot prove

• Sellability, transfer restrictions, actual tax rates, gas, MEV or future execution prices. Zero assumed tax is not a claim of zero actual tax.
• Whether liquidity is locked or can be removed. Proportional removal is a hypothetical stress.
• V3, Infinity, stable-swap curves, other factories or multi-hop routing.
• Market-wide coverage, safety, investment merit or demand for this product. The pilot covers two pools.

Evidence and baselines

Two real pools were initially tested at 0.1%, 1% and 10% of input reserves. All six outputs matched the deployed router exactly. This validates the reserve calculation; it does not demonstrate profitable trading or product-market fit. A human retention pilot has not been performed.

V2 math source · Swap fees · Metadata API

Public data API

GET /api/v1/snapshots?page=1&limit=50 returns { items, page, per_page, total, next }. Limits are capped at 50. /api/v1/stats returns current pilot observations with the same envelope. Missing sources produce an explicit error, never fabricated reserves.

Privacy

No wallet connection is required. Bag sizes and assumptions stay in your browser until you copy a link. Shared links include those values. Avoid sharing a drill if you want its hypothetical position size to remain private.