Pyth and SEDA
Oracle data pipeline (Pyth -> SEDA -> Relayer -> Hyperliquid)
Dreamcash markets use an oracle pipeline that is designed to be independent, verifiable, and hard to manipulate. No single party controls the full path from raw price sourcing to on-chain delivery.
At a high level:
Pyth provides the underlying market price feed, sourced from multiple independent publishers.
SEDA applies Dreamcash session logic (Regular Hours vs Off Hours) and produces a single oracle output stream.
An independent relayer reads the SEDA output and pushes oracle updates to the on-chain market using an authorized oracle address.
Why this design matters
This separation provides three key properties:
Independence: sourcing (Pyth), logic (SEDA), and delivery (relayer) are operated as separate components.
Verifiability: each stage can be inspected independently, including the external feed used, the logic used to transform it, and the on-chain transactions that update the market.
Consistency: session rules are applied programmatically rather than manually.
Components
Pyth
A price-feed network that aggregates data from multiple independent publishers
Provides the external reference feed during Regular Hours
SEDA
Programmable oracle infrastructure
Applies Dreamcash rules for Regular Hours vs Off Hours and outputs a single oracle stream
Relayer
An independent process operated separately from the UI and from data providers
Reads the SEDA stream and submits signed oracle update transactions with redundancies
Pyth
Pyth is a network that publishes price feeds for many assets. The important point for Dreamcash users is how the prices are produced:
Multiple independent publishers contribute data.
The feed is aggregated into a single published price.
The feed can be inspected publicly (for example, to see which feed is being used and its update behavior).
For Dreamcash equity indices, the Pyth feed is the external reference used during Regular Hours.
Example feed used for US500:
https://insights.pyth.network/price-feeds/Equity.US.EMH6%2FUSD
SEDA
SEDA is oracle infrastructure that lets you define an oracle as a program. Instead of consuming a single raw feed directly, you can specify logic like:
which source(s) to read (for example, a Pyth feed),
how to process or smooth the data,
how to switch behavior depending on conditions.
Dreamcash uses SEDA to apply the session logic described in Trading -> Oracles and Pricing:
During Regular Hours, the oracle output is derived from the external Pyth feed (and any required transformations).
During Off Hours, the oracle output follows Dreamcash internal-session logic (for example, using smoothed internal measurements).
Because the logic is encoded as a program, it is consistent and easier to audit than a manual process.
What the relayer does
The relayer is the delivery layer:
It watches the SEDA oracle output stream.
When an update is required, it submits an on-chain transaction that updates the market’s oracle price.
Those transactions are signed by an authorized oracle address, and they are publicly visible on-chain.
The relayer is run independently from the Dreamcash frontend. If the website is unavailable, the oracle pipeline can still function.
Last updated