Magicblock logo

Magicblock

Organization
sendaifun
magicblock

Complete guide for MagicBlock Ephemeral Rollups - high-performance Solana execution with sub-10ms latency, gasless transactions, and Solana Plugins. Use when building real-time games, high-frequency trading, or any application requiring ultra-low latency on Solana.

Overview

Publishersendaifun
Repositoryskills
Skill namemagicblock
Stars
128
Forks
81
Bundled files
6
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 6 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by sendaifun on GitHub. Read the source before you install it.

Installation

Install the Magicblock AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/sendaifun/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/skills/magicblock .claude/skills/magicblock
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Magicblock in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Magicblock on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Magicblock is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

MagicBlock Ephemeral Rollups Guide

A comprehensive guide for building high-performance Solana applications with MagicBlock Ephemeral Rollups - enabling sub-10ms latency and gasless transactions.

Overview

MagicBlock Ephemeral Rollups (ER) are specialized SVM runtimes that enhance Solana with:

  • Sub-10ms latency (vs ~400ms on base Solana)
  • Gasless transactions for seamless UX
  • Full composability with existing Solana programs
  • Horizontal scaling via on-demand rollups

Architecture

┌─────────────────────────────────────────────────────────────┐
│                      Your Application                        │
├─────────────────────────────────────────────────────────────┤
│  Base Layer (Solana)          │  Ephemeral Rollup (ER)      │
│  - Initialize accounts        │  - Execute operations       │
│  - Delegate accounts          │  - Process at ~10-50ms      │
│  - Final state commits        │  - Zero gas fees            │
│  - ~400ms finality            │  - Commit state to Solana   │
└─────────────────────────────────────────────────────────────┘

Core Flow

  1. Initialize - Create accounts on Solana base layer
  2. Delegate - Transfer account ownership to delegation program
  3. Execute - Run fast operations on Ephemeral Rollup
  4. Commit - Sync state back to base layer
  5. Undelegate - Return ownership to your program

Prerequisites

bash
# Required versions
Solana: 2.3.13
Rust: 1.85.0
Anchor: 0.32.1
Node: 24.10.0

# Install Anchor (if needed)
cargo install --git https://github.com/coral-xyz/anchor anchor-cli

Quick Start

1. Add Dependencies (Cargo.toml)

toml
[dependencies]
anchor-lang = "0.32.1"
ephemeral-rollups-sdk = { version = "0.6.5", features = ["anchor", "disable-realloc"] }

2. Program Setup (lib.rs)

rust
use anchor_lang::prelude::*;
use ephemeral_rollups_sdk::anchor::{delegate_account, commit_accounts, ephemeral};
use ephemeral_rollups_sdk::cpi::DelegationProgram;

declare_id!("YourProgramId111111111111111111111111111111");

#[ephemeral]  // Required: enables ER support
#[program]
pub mod my_program {
    use super::*;

    pub fn initialize(ctx: Context<Initialize>) -> Result<()> {
        ctx.accounts.state.value = 0;
        Ok(())
    }

    #[delegate]  // Auto-injects delegation accounts
    pub fn delegate(ctx: Context<Delegate>) -> Result<()> {
        Ok(())
    }

    pub fn increment(ctx: Context<Update>) -> Result<()> {
        ctx.accounts.state.value += 1;
        Ok(())
    }

    #[commit]  // Auto-injects commit accounts
    pub fn undelegate(ctx: Context<Undelegate>) -> Result<()> {
        Ok(())
    }
}

3. TypeScript Client Setup

typescript
import { Connection, PublicKey } from "@solana/web3.js";
import { AnchorProvider, Program } from "@coral-xyz/anchor";
import { DELEGATION_PROGRAM_ID } from "@magicblock-labs/ephemeral-rollups-sdk";

// CRITICAL: Separate connections for each layer
const baseConnection = new Connection("https://api.devnet.solana.com");
const erConnection = new Connection("https://devnet.magicblock.app");

// Create providers
const baseProvider = new AnchorProvider(baseConnection, wallet, { commitment: "confirmed" });
const erProvider = new AnchorProvider(erConnection, wallet, {
  commitment: "confirmed",
  skipPreflight: true,  // Required for ER
});

// Check delegation status
async function isDelegated(pubkey: PublicKey): Promise<boolean> {
  const info = await baseConnection.getAccountInfo(pubkey);
  return info?.owner.equals(DELEGATION_PROGRAM_ID) ?? false;
}

Key Concepts

Delegation

Delegation transfers PDA ownership to the delegation program, allowing the Ephemeral Validator to process transactions.

rust
#[derive(Accounts)]
pub struct Delegate<'info> {
    #[account(mut)]
    pub payer: Signer<'info>,
    /// CHECK: Will be delegated
    #[account(mut, del)]  // 'del' marks for delegation
    pub state: AccountInfo<'info>,
    pub delegation_program: Program<'info, DelegationProgram>,
}

Commit

Commits update PDA state from ER to base layer without undelegating.

rust
use ephemeral_rollups_sdk::anchor::commit_accounts;

pub fn commit(ctx: Context<Commit>) -> Result<()> {
    commit_accounts(
        &ctx.accounts.payer,
        vec![&ctx.accounts.state.to_account_info()],
        &ctx.accounts.magic_context,
        &ctx.accounts.magic_program,
    )?;
    Ok(())
}

Undelegation

Returns PDA ownership to your program while committing final state.

rust
#[commit]  // Handles commit + undelegate
pub fn undelegate(ctx: Context<Undelegate>) -> Result<()> {
    Ok(())
}

ER Validators (Devnet)

RegionValidator Identity
AsiaMAS1Dt9qreoRMQ14YQuhg8UTZMMzDdKhmkZMECCzk57
EUMEUGGrYPxKk17hCr7wpT6s8dtNokZj5U2L57vjYMS8e
USMUS3hc9TCw4cGC12vHNoYcCGzJG1txjgQLZWVoeNHNd
TEEFnE6VJT5QNZdedZPnCoLsARgBwoE6DeJNjBs2H1gySXA

Magic Router (auto-selects best): https://devnet-router.magicblock.app

Critical Rules

DO:

  • Maintain separate connections for base layer and ER
  • Use skipPreflight: true for all ER transactions
  • Verify delegation status before sending to ER
  • Use AccountInfo for delegated accounts in Rust
  • Match PDA seeds exactly between Rust and TypeScript

DON'T:

  • Send delegated account operations to base layer
  • Mix base layer and ER operations in single transaction
  • Assume account ownership without checking
  • Skip commitment verification before base layer reads

Products

ProductDescription
Ephemeral Rollup (ER)High-performance, gasless transactions
Private ER (PER)Privacy-preserving computation with Intel TDX
VRFVerifiable random function for on-chain randomness
BOLT FrameworkECS architecture for fully on-chain games
Solana PluginsApp-specific extensions for enhanced capabilities

Solana Plugins (New)

Solana Plugins are modular capabilities that can be added to your dApp to extend what's possible on Solana. Think of them as your custom toolkit: plug in what you need, when you need it.

Available Plugins

PluginDescriptionUse Cases
Verifiable Randomness (VRF)Provably fair on-chain randomnessGames, lotteries, NFT drops
Real-Time Price FeedsUp-to-the-millisecond market dataDEXs, trading bots, DeFi
AI OraclesCall AI models directly from smart contractsDynamic NFTs, AI agents

Using VRF Plugin

typescript
import { requestRandomness, getRandomnessResult } from "@magicblock-labs/vrf-sdk";

// Request randomness
const requestTx = await requestRandomness({
  payer: wallet.publicKey,
  seed: Buffer.from("my_game_seed"),
});

// Get result after confirmation
const result = await getRandomnessResult(requestId);
console.log("Random value:", result.randomness);

Privacy with Intel TDX

MagicBlock enables privacy in any Solana program state account through Ephemeral Rollups running in Trusted Execution Environments (TEE) on Intel TDX. This allows:

  • Private computation without exposing state
  • Verifiable execution guarantees
  • Selective disclosure of results

Resources

Skill Structure

magicblock/
├── SKILL.md                          # This file
├── resources/
│   ├── api-reference.md              # Complete API reference
│   └── program-ids.md                # All program IDs and constants
├── examples/
│   ├── anchor-counter/README.md      # Basic counter with delegation
│   ├── delegation-flow/README.md     # Full delegation lifecycle
│   ├── vrf-randomness/README.md      # VRF integration
│   └── crank-automation/README.md    # Scheduled tasks
├── templates/
│   ├── program-template.rs           # Rust program starter
│   └── client-template.ts            # TypeScript client starter
└── docs/
    ├── advanced-patterns.md          # Complex patterns
    └── troubleshooting.md            # Common issues

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Magicblock AI skill do?

Complete guide for MagicBlock Ephemeral Rollups - high-performance Solana execution with sub-10ms latency, gasless transactions, and Solana Plugins. Use when building real-time games, high-frequency trading, or any application requiring ultra-low latency on Solana.

Why use Magicblock on TypingMind?

Because you install it once and use it with any model. Magicblock is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Magicblock in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/sendaifun/skills/tree/main/skills/magicblock. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Magicblock?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Magicblock?

As many as you like. As long as a model supports skills, you can use Magicblock with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Magicblock AI skill free?

Yes. It is published on GitHub by sendaifun under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇