> ## Documentation Index
> Fetch the complete documentation index at: https://docs.hadron.fi/llms.txt
> Use this file to discover all available pages before exploring further.

# Spread Trigger Rules

> How to use spread triggers, how not to, and what happens if you cross the line

Spread triggers let you price flow you can name: a bot program, an arbitrage wallet, a partner
frontend. They are not a way to widen your quote for everyone who routes through Jupiter, Titan or
OKX. This page is the line between the two. If a script or an agent manages your trigger table,
give it this page.

## How to use it

Use a trigger when you can finish the sentence "this row exists because of this specific
counterparty". Good targets:

* **A wallet you have observed.** A signer that arbitrages your pool, or a sniper. Ban it with
  10,000 bps or more under the wildcard program. Banning a wallet seconds after its first trade
  is fine. The target is the wallet.
* **A bot program on your pool.** Key the row on your own config, with the bot's program id as the
  program filter. It fires only when that program calls your pool.

## How NOT to use it

Do not use a trigger to add spread to all of your routed flow. Routers quote your pool without
knowing which triggers will fire, so a row that fires on every swap is a hidden tax: the user is
quoted one price and filled at a worse one. Rows that do this:

* **Your pool itself.** Your config, vaults, curve meta, spread config, pool fee config, mints or fee
  recipient token accounts, under the wildcard or under a router program. Every swap on your pool
  carries these, and routers copy every one of them into their own instruction.
* **All routers.** Jupiter, Titan, OKX or DFlow program ids, their event authority PDAs, the jup.ag
  frontend signer `sighWH8KaiT7QhtV4w29ReVF8kG6D5yG3EQP1KYyGVF`, the Jupiter Ultra relayer
  `AgmLJBMDCqWynYnQiPCuj9ewsNNsBJXyzoUhD9LJzN51` or Titan's authority
  `D5YqVMoSxnqeZAKAUUE1Dm3bmjtdxQ5DCF356ozqN9cM`, under any program filter.
* **All competing venues.** Other venues' program ids or pools under the wildcard. A retail route
  that hops through that venue pays it too, and no router quotes it.
* **Accounts no swap can avoid.** The Token programs, the Associated Token program, the wSOL, USDC
  or USDT mints, the global `FeeConfig`, the Clock or Instructions sysvars.
* **Tips and markers.** Jito tip accounts, `jitodontfront` markers, the Memo program or Jupiter
  referral. These ride on a large share of ordinary retail flow.
* **`NEGATE` under the wildcard or a router.** It fires when the account is absent, which is almost
  every swap, whatever account it names.

Keying on a vault instead of the config, or on your wSOL fee account, is the same thing written
differently. Every account on your pool's own list counts.

## Some examples of what's okay

| Row | Reads as |
| :- | :- |
| `Bsc1BN…` wallet, any program, +20,000 bps | "This sniper traded my pool once. It is done." |
| My config, program `SiVKAT…`, +200 bps | "This arb program pays more when it calls my pool." |
| Partner signer, `SIGNER`, -20 bps | "Users of this frontend get a tighter price." |
| `SysvarRecentB1ockHashes…`, System program, +300 bps | "Durable-nonce transactions are bots." |

## Some examples of what's not okay

| Row | Why |
| :- | :- |
| My config, any program, +500 bps | Every swap on the pool pays it |
| My SOL vault, Jupiter program, +300 bps | Same row, keyed on a vault instead |
| `JUP6Lk…` (Jupiter), any program, +290 bps | Every Jupiter swap pays it, on any pool |
| `sighWH8K…` (jup.ag signer), System program, +50 bps | Every jup.ag frontend swap pays it |
| Instructions sysvar, any program, +990 bps | Every swap carries the sysvar |
| Competing venue program ids, any program, +15 bps each | Every route that hops through that venue pays it |
| `jitodontfront…`, any program, +60 bps | Jupiter Ultra and Titan attach it to most flow |
| Any account, `NEGATE`, any program | Fires on nearly every swap |

We may flag other tables beyond these examples. The goal is to prevent retail abuse.

## Why this matters

Routers see your base spread and quote it correctly. They cannot see triggers. A table that taxes
all routed flow is paid out of users' slippage budgets by people who never agreed to it, and the
router learns that your fills under-deliver and routes less to you. It also puts Hadron's standing
with every router at risk, which is why we enforce it for every pool, not just yours.

The cost is real even at small sizes. On one SOL/USDC pool a wildcard list of 32 venue programs at
10 bps each fired on 33% of retail fills, for 10 to 50 bps that no router quoted.

## Consequences

We monitor every live spread config and will reach out to flag a problem, barring which we will
pause the pool. A paused pool drops out of Jupiter, Titan and OKX until the rows are fixed.

## Questions?

Reach out on Telegram to Adam at [@adamgpeters](https://t.me/adamgpeters) and Srinath at
[@srhhadron](https://t.me/srhhadron). We would rather talk through a table before it goes live than
pause a pool after.

See [Routing and Flow Segmentation](https://docs.hadron.fi/makers/routing-and-flow.md) for how routers quote your pool and
[Segmenting flow](https://docs.hadron.fi/sdk/segmenting-flow.md) for the SDK calls.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.