v0.1.29

feat: Add official website and configure /app base path for product frontend

• Build DeepDiagram official website with Next.js 15, next-intl (zh/en),
  MDX blog system, and full SEO (sitemap, robots, JSON-LD, OG images)

• Add pages: Home, Features, Pricing, Blog, Docs, About, Changelog

• Write documentation (Getting Started, Deployment Guide, API Reference)
  in both Chinese and English based on actual project source code

• Add feature demo images and product demo GIF to website

• Configure frontend SPA to serve under /app base path (Vite base + nginx alias)

• Add website Docker image build to CI/CD pipeline (GitHub Actions)

• Add website service to docker-compose.yml

• Update all product links from deepd.cturing.cn to deepd.cturing.cn/app

• Update SITE_URL to https://deepd.cturing.cn across sitemap, robots,
  metadata, and JSON-LD structured data

v0.1.28

feat: Add official website and configure /app base path for product frontend

• Build DeepDiagram official website with Next.js 15, next-intl (zh/en),
  MDX blog system, and full SEO (sitemap, robots, JSON-LD, OG images)

• Add pages: Home, Features, Pricing, Blog, Docs, About, Changelog

• Write documentation (Getting Started, Deployment Guide, API Reference)
  in both Chinese and English based on actual project source code

• Add feature demo images and product demo GIF to website

• Configure frontend SPA to serve under /app base path (Vite base + nginx alias)

• Add website Docker image build to CI/CD pipeline (GitHub Actions)

• Add website service to docker-compose.yml

• Update all product links from deepd.cturing.cn to deepd.cturing.cn/app

• Update SITE_URL to https://deepd.cturing.cn across sitemap, robots,
  metadata, and JSON-LD structured data

v0.1.27

fix: nginx config

v0.1.26

fix: 输入法组合状态下按回车会确认输入的英文字母，而不是直接发送消息。只有在非组合状态下按回车才会发送

v0.1.25

fix: DrawioAgent render

v0.1.24

sitemap.xml

v0.1.23

修复 general agent 聊天刷新后数据丢失的 bug

v0.1.22

feat: Upgrade AntV Infographic to 0.2.11 and refactor agent output format

• Upgrade @antv/infographic from 0.2.6 to 0.2.11
• Refactor all agents to use XML tag output format (<design_concept>/)
instead of JSON, simplifying parsing and supporting multi-line content
• Add StreamingTagParser to replace StreamingJsonParser for robust tag parsing
• Enhance Flow and Draw.io agent prompts with detailed syntax examples to
  reduce parsing errors
• Implement two-phase pipeline for Infographic agent (template selection +
  code generation)
• Update README documentation with new architecture diagrams

v0.1.21

feat: Add model connection testing before saving custom LLM configurations

• Add /api/test-model endpoint to validate model configurations via a simple API call
• Frontend tests connection before allowing save, showing success/error feedback
• Display success notification for 1.5s after model is successfully added/updated
• Clear test results when form fields change

Co-Authored-By: Claude Opus 4.5 noreply@anthropic.com

v0.1.20

架构简化， 移除 tool call 的逻辑， 避免因模型能力不调用工具从而不生图。