Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Repository files navigation

HelloJourney

A full-stack AI travel planning workspace for turning a destination, budget, and preferences into an actionable itinerary.

Repository · GPL-2.0 License

HelloJourney 是一个可验证、可编辑、可局部重新规划、可撤销的 AI Travel Workspace。系统通过原生 Tool Calling 调用地图与天气能力,用结构化输出和 Review Agent 约束行程质量,并提供 Web 端(React)和 微信小程序(UniApp)体验。

HelloJourney combines a React web client, a Spring Boot service, and a UniApp / WeChat mini-program client. It is designed as a real product workflow: model output is enriched with map data, weather, travel content, structured POIs, and live task progress.

Why it is interesting

  • AI itinerary planning with multi-provider LLM support and context-aware follow-up chat.
  • Agentic pipeline: native tool calling, structured JSON output, and a Review Agent that blocks invalid plans.
  • Real integrations for Tencent Maps, AMap, Google Maps, weather, POI search, geocoding, and route planning — with automatic provider fallback.
  • Verified attraction photos via AMap POI identity matching; deterministic placeholder cards when nothing is verified.
  • Editable itinerary workspace with undo/redo, cross-day moves, and AI-proposed partial replan change sets you can apply or reject.
  • Knowledge graph view built with ECharts for exploring destinations and relationships.
  • Interactive itinerary map (dark 3D AMap) with numbered markers, info windows, and route fallbacks.
  • zh / en / ja i18n, itinerary PNG export, reservation reminders, and a read-only settings page that never exposes secrets.
  • Async generation pipeline with WebSocket progress updates, retryable sections, and history.
  • Multi-client delivery across React Web, UniApp, and WeChat mini-program.
  • Mock-first development with MSW, plus backend tests with JUnit 5 and Mockito.

Stack

Layer Technologies
Web React 19 · TypeScript · Vite · Ant Design · ECharts · i18next
Service Java 17 · Spring Boot 3 · WebSocket · OkHttp
Mobile UniApp · Vue 3 · TypeScript
AI / Data DeepSeek · OpenAI-compatible providers · Tencent Maps · AMap · Google Maps
Quality MSW · JUnit 5 · Mockito · Vitest · ESLint

Quick start

Requirements: JDK 17+, Maven 3.8+, Node.js 20+, and npm 10+.

git clone https://github.com/kdjfs/HelloJourney.git
cd HelloJourney

Start the service:

cd backend
mvn spring-boot:run -Dspring-boot.run.profiles=dev

Start the web client in another terminal:

cd frontend
npm ci
npm run dev

The web client runs at http://localhost:5173; the service defaults to http://localhost:8000. Configure provider keys and map credentials in backend/src/main/resources/application.yml. Keep secrets local and out of Git.

For a frontend-only walkthrough, use the repository's MSW mock mode with VITE_USE_MOCK=true.

Project shape

backend/   Spring Boot API, AI orchestration, integrations, WebSocket progress
frontend/  React web application and mock-first API layer
uniapp/    Vue 3 cross-platform / WeChat mini-program client

Status

This is an active engineering project. The current README documents implemented capabilities and remaining roadmap items; external API credentials are intentionally required for live provider flows.

About

Full-stack AI travel planning platform with React, Spring Boot, WebSocket, maps, knowledge graph, and UniApp clients.

Topics

Resources

Stars

5 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages