hyperstack

Original:🇺🇸 English
Translated

Build with Hyperstack, real-time Solana data streaming. Covers consuming existing streams and building custom stacks. Use when the user mentions Hyperstack, real-time Solana data, on-chain streaming, ORE mining, or wants to consume/build data stacks.

11installs
Added on

NPX Install

npx skill4agent add usehyperstack/skills hyperstack

Tags

Translated version includes tags in frontmatter

Hyperstack

Hyperstack provides real-time streaming data from Solana programs via WebSocket. Think of it as a typed, structured pub/sub system for on-chain state.
Key concepts:
  • Stack: A deployed streaming server that watches Solana programs and produces structured entity updates
  • Entity: A domain object (e.g.
    OreRound
    ,
    OreMiner
    ) with typed fields organized in sections
  • View: A lens on an entity, either
    state
    (keyed lookup) or
    list
    (collection). Custom views can add sorting and filtering.
  • SDK: TypeScript, React, or Rust client libraries that connect to a stack's WebSocket and provide typed access to views

CLI Setup

The Hyperstack CLI has different binary names depending on how it was installed:
  • cargo install:
    hs
  • npm install -g:
    hyperstack-cli
Before running any CLI commands, detect which binary is available:
bash
# Check which CLI is available (prefer hs)
if command -v hs &> /dev/null; then
  HS_CLI="hs"
elif command -v hyperstack-cli &> /dev/null; then
  HS_CLI="hyperstack-cli"
else
  echo "Hyperstack CLI not found. Install with: cargo install hyperstack-cli (recommended) or npm install -g hyperstack-cli"
  exit 1
fi
Then use
$HS_CLI
for all commands, or just use
hs
directly if you confirmed it's available.

Quick Start

bash
# Scaffold a new app from a template
npx hyperstack-cli create my-app

# Or install the CLI globally (cargo is preferred)
cargo install hyperstack-cli  # Installs as 'hs'
# OR
npm install -g hyperstack-cli  # Installs as 'hyperstack-cli'

Discovering Available Stacks and Schemas

CRITICAL: Always run the explore command before writing Hyperstack code. Never guess entity names, field paths, or types.
bash
# List all available stacks (use 'hs' or 'hyperstack-cli' depending on your install)
hs explore --json

# Get entities and views for a specific stack
hs explore <stack-name> --json

# Get detailed field info for a specific entity
hs explore <stack-name> <EntityName> --json
Note: All CLI examples use
hs
. If you installed via npm, substitute
hyperstack-cli
for
hs
.
The
--json
flag gives machine-readable output with exact field paths, types, and view definitions.

Which Skill Do You Need?

Consuming an existing stack (connecting to streams, subscribing to data, building UIs): -> See the
hyperstack-consume
skill
Building your own stack (writing Rust DSL, defining entities from program IDLs, deploying): -> See the
hyperstack-build
skill

Prepackaged Stacks

Run
hs explore
to see what's available. Current public stacks include:
  • ore, ORE mining rounds, miners, treasury state
Install the prepackaged SDK types:
npm install hyperstack-stacks