Quick Start
Method 1: Desktop App (Recommended)
Download the installer → Install → Launch → Start using
| Platform | Download |
|---|---|
| Windows | HippoBuddy Setup |
| macOS (Intel) | HippoBuddy.dmg |
| macOS (Apple Silicon) | HippoBuddy-arm64.dmg |
| Linux (AppImage) | HippoBuddy.AppImage |
Method 2: Run from Source
# 1. Build the Java backend
mvn package -DskipTests
# 2a. Launch desktop app (Electron)
cd electron && npm install && npm start
# 2b. Or launch web-only (without Electron)
mvn exec:java -Dexec.mainClass="com.example.agent.WebApplication"
Configuration
When running from source, the app will create config.yaml from config.yaml.example on first launch. Edit the LLM configuration:
llm:
api_key: ${DEEPSEEK_API_KEY:-your-api-key-here}
model: deepseek-v4-flash
base_url: https://api.deepseek.com
Supports DeepSeek / Claude / GPT / Ollama. See config.yaml.example in the project for the full configuration reference.
Project Structure
src/main/java/com/example/agent/
├── WebApplication.java Web entry point
├── DesktopApplication.java Desktop entry point
├── core/ DI, event bus, security interceptor
├── llm/ LLM clients (OpenAI, Claude, Ollama...)
├── tools/ Built-in tools (20+)
├── execute/ Agent conversation loop
├── orchestrator/ Task orchestration (DAG)
├── subagent/ Multi-agent system
├── mcp/ MCP protocol integration
├── memory/ Long-term memory
├── session/ Session storage & transcripts
├── web/ HTTP handlers & SSE streaming
├── prompt/ Prompt library & management
├── domain/ Rules, skills, content truncation
└── config/ Configuration center