# Hooks

> The extension point pools use to add rules, and what it does not promise.

import { HookLedgerAnimation } from "../../../components/docs/hook-ledger-animation";
import { HookPointsAnimation } from "../../../components/docs/hook-points-animation";

## Extension Point: Hooks

The protocol's core logic — pricing, settlement, liquidity management — behaves identically across every pool. In practice, though, individual pools often need their own rules: dynamic fees that rise only under heavy volume, access restrictions that permit trading only under certain conditions, or liquidity strategies that automatically rebalance around each settlement. **Hooks** are the extension point that lets this kind of pool-specific logic attach as a separate contract, without touching the protocol core.

A hook can intervene before and after each of the five actions a pool supports — initialization, adding liquidity, removing liquidity, swapping, and donating. Each action has a "before" and an "after" point, for ten intervention points in total. Which of these a pool uses is chosen when the pool is created; any point not chosen is simply never invoked.

<HookPointsAnimation />

Take the swap points alone as an example: a pre-trade intervention can alter the amount that actually settles, while a post-trade intervention can require additional asset movement layered on top of an already-finalized result.

### Permission Is Fixed When the Pool Is Created

What a hook is allowed to do is fixed the moment its pool is created — the pool is only created if the hook's self-declared capability list matches the pool's configuration exactly, and that scope never changes afterward.

### A Hook's Settlement Is Recorded Separately

Any additional asset movement a hook generates is always kept separate from the trader's own settlement, and passes through the same transaction ledger check.

<HookLedgerAnimation />

This check only guarantees that a hook cannot break accounting integrity — not that its judgment is always fair. What logic a hook actually runs within its declared scope is outside the protocol's concern.

## Properties and Trade-offs

- **Leaving the range stops yield immediately.** A position earns nothing from subsequent trades until price re-enters its range — not a defect, but an inherent property of concentrating liquidity in a range.
- **A range with a single liquidity holder can have its cumulative fee counter artificially inflated.** If the sole LP donates to itself, the cumulative value can rise abnormally.
- **The protocol's fee cut is not a fixed value.** A portion of LP fees can be split off at a governance-set rate, which may change over time.
- **Using a pool with a hook means trusting that hook too.** The protocol only prevents a hook from exceeding its declared scope — it makes no guarantee about the logic inside that scope.