Developer Alpha v0

HOL Desktop

HOL Desktop is a cross-platform app for interacting with AI agents on Hedera. Connect MCP servers, execute transactions, and manage hashgraph operations through natural conversation.

View on GitHubJoin Community
Desktop App

Cross-platform Electron app for Windows, macOS, and Linux

MCP Support

Model Context Protocol for AI-system integration

Extensible

Plugin architecture for custom tools and workflows

Conversational Agent Desktop App

Two Ways to Build

Choose between our Electron desktop app with a full GUI or the CLI built with Ink for terminal enthusiasts.

Conversational Agent Desktop Chat
HOL Desktop

Electron desktop app with modern React UI. Chat with AI agents, approve transactions, connect MCP servers for file access and external tools.

Key Features
  • Modern chat interface with transaction approval cards
  • File attachments and multi-modal support
  • HCS-10 profile integration with AI agent display
  • Real-time connection status and network indicators
Installation
install.sh
# Clone and build from source:
git clone https://github.com/hashgraph-online/desktop
cd desktop
pnpm install
# For development
pnpm dev
# For production build
pnpm build

Developer Tools & Resources

Everything you need to build with AI agents on Hedera

hol-desktop-tools.dev
6 tools available
tool directory

HOL Desktop

Full-featured Electron app with chat interface, transaction approvals, and MCP server management. Cross-platform support for Windows, macOS, and Linux.

QUICK INSTALL

Copy
$git clone https://github.com/hashgraph-online/desktop

QUICK START

# Clone and run locally
git clone https://github.com/hashgraph-online/desktop
cd desktop
pnpm install
pnpm dev


# For production build
pnpm build
View on GitHubDocumentation

Technical Architecture_

Production-ready AI agent infrastructure built on LangChain, Hedera, and MCP

architecture-explorer.sh
4 tools available
tool directory

LangChain Architecture

Built on LangChain with OpenAI function calling, ContentAware AgentExecutor, and smart memory management for context-aware conversations.

QUICK START

const agent = new ConversationalAgent({
  accountId: process.env.HEDERA_ACCOUNT_ID!,
  privateKey: process.env.HEDERA_PRIVATE_KEY!,
  network: 'testnet',
  openAIApiKey: process.env.OPENAI_API_KEY!,
  openAIModelName: 'gpt-4o',
  verbose: true
});


// Smart memory with entity tracking
const response = await agent.processMessage(
  'Send 1 HBAR to alice.eth'
);


// Features:
// - OpenAI function calling
// - ContentAware AgentExecutor  
// - Smart memory management
// - Entity resolution system
View on GitHubDocumentation
// LANGCHAIN_POWERED

AI-First Design

Built on LangChain with OpenAI/Anthropic models and smart memory management

// HEDERA_NATIVE

Hashgraph Ready

Full Hedera integration with HCS standards and hedera-agent-kit v2

// MCP_ENABLED

Extensible

Model Context Protocol for external systems and custom tool integration

quick-start.sh
npm install @hashgraphonline/conversational-agent
✓ Installing conversational agent SDK...
✓ Setting up LangChain architecture...
✓ Configuring Hedera plugins...
✓ Initializing MCP servers...
// Ready to build AI agents on Hedera

Don't Miss Out!

Be the first to know about workshops, deadlines, and get exclusive resources to boost your hackathon success.

Workshop alerts
Key deadlines
Exclusive resources
Early announcements