Roadmap v0.7 #200
Description
What have we done previously?
v0.7
We made 301 commits, changed 891 files, updating, refactoring over 48,100+ 17,452- lines of code, having 23 contributors in total during the development from v0.6.1 to v0.7.
Milestones:
- We reached 200 Discord 👾 members on June 16, 2025
- We reached 1000 stars 🌟 on June 21, 2025
- We announced on Hacker News
- We announced on Product Hunt
- We are trending on GitHub
#1🏆 on July 17, 2025 - We reached 1500 stars 🌟 on July 19, 2025
New packages born between this release:
Big shout out to @sumimakito, she did so many fantastic things... I can't even count them all...
@proj-airi/chromatic(by @sumimakito)@proj-airi/unocss-preset-chromatic(by @sumimakito)@moeru-ai/jem(by @LemonNekoGH), unified model catalogclustr(by @sumimakito)@proj-airi/drizzle-orm-browser(by @nekomeowww)
Side projects born between this release:
-
HuggingFace Inspector (https://github.com/moeru-ai/hf-inspector)
-
More candle examples about whisper & VAD, candle, burn, and ort
-
Features Requests
- feature request: prompt user with API key and Base URL setup on first launch #254
- Feature Proposal: Live2D Model Scaling and Positioning Controls #265
- feature request: include error messages when configuring Speech Providers #280
- feature request: redesign Voice Configuration error container #294
- feature request: use avif instead of png/jpg #298
-
Bugs
- "Proxy object could not be cloned." when sending more than 1 message #192
- stage-web:
<description>tag appears in<body>#230 - Model disappeared cross desktops ([macOS] Model Disappears When Switching Desktops (Spaces) #222), fixed by 223
- Feature Proposal: Live2D Model Scaling and Positioning Controls #265
- Feature Proposal: Live2D Model Scaling and Positioning Controls #265
- Grok(x.ai) API support? #241
- Is Airi Docs' site broken? #273
- bug: stage-web imports
virtual:pwa-register#281 - bug: Error fetching models for deepseek: TypeError: Failed to construct 'URL': Invalid base URL #291
-
UI
- New Landing Page (5e8b2b8, feat(docs): migrate to vitepress with @unovue/reka-ui's theme #224)
- New Documentation Page (5e8b2b8, feat(docs): migrate to vitepress with @unovue/reka-ui's theme #224)
- Onboarding screen (thanks to @Me1td0wn76, feat(stage-web): launch setup #259)
- Components
- Text slice / Grapheme splitter by @sumimakito's clustr (7a04fd6)
- refactor(stage-ui): replace grapheme cluster reader from clustr #308
- Poppin'text for subtitles (db38c22)
- Chat bubble (c449df5, feat(stage-ui): Chat Bubble with minimalism theme #267)
- Fade animation support (b56b8bf)
- Chat bubble (c449df5, feat(stage-ui): Chat Bubble with minimalism theme #267)
-
<Progress />(thanks to @Menci 2cb602a) -
<TimeSeriesChart />(c855047) -
<ThresholdMeter />(5573f2a) - ` (c00e112)
-
<Progress />(2cb602a) -
<FieldSelect />(d0d782f) -
<Alert />(thanks @typed-sigterm, eeadf3b, refactor(stage-ui):<Alert> <ErrorContainer>#295) -
<ErrorContainer />(thanks @typed-sigterm, eeadf3b, refactor(stage-ui):<Alert> <ErrorContainer>#295) - New sidebar nav design
- Toaster
- Prompt to update for user when new version is available
- Transcription playground
- Speech playground
- Lorebook editor
- Text slice / Grapheme splitter by @sumimakito's clustr (7a04fd6)
- Providers
- Azure AI Foundry (f471bec, feat: Add Azure AI Foundry provider for consciousness #272, thanks to @Inokinoki)
- OpenAI Audio Services
- Better TTS input chunking system (42026d1, feat(stage-ui): better tts input chunking #306, thanks to @sumimakito)
- Tool discovery and validation (9aa5fe9, fix(stage-*): passed
toolswithout check & discovery for provider #296, thanks to @LemonNekoGH, and me @nekomeowww)
- Theme
- Web
- Dramatically improved mobile interfaces (0db4fa3)
- Mobile-optimized chat UI (4841fa9)
- Mobile range dragging fixes (a8c668b)
- fix: pwa service worker registering #299
- KeepAlive stage implementation (3806d4b)
- Mobile sending button (1ae7cc3)
- Model offset and blur effects (7478b5a)
- Interactive area improvements (4ffe2e4)
- Desktop
- Hide on hover behavior (8052462, thanks to @LemonNekoGH, @sumimakito)
- Windows build compatibility (87c799c)
- Window state (206e5e5, aa8f914, 8ee87f9, thanks to @sumimakito)
- All workspace visibility (ae98a39, feat(stage-tamagotchi): allow visible on all workspace #223, @LemonNekoGH)
- WindowLink component system (94048b0)
- Tray & Menu (befbb18, thanks to @LemonNekoGH)
- Resource Island (2c93bf1)
- WindowRouterLink plugin (5d5c7c5)
- Audio transcription plugin (thanks to @luoling8192)
- Audio VAD plugin (@luoling8192)
- Inline plugin system (ad0d4e9, refactor(stage-tamagotchi): make all small components into inline plugins #248)
- Tooltip when loading (from @nekomeowww)
- Prompt user to set the API key and base URL
- Provider configuration
- Kokoro TTS
- Spark TTS
- GPT-VITS
- Speaches.ai (from @kwaa)
- Dubbingx (quite slow inference, not urgent to support)
- Custom LLM model serving
- Fine tuning dataset helper editor & viewer
- Serving web ui
-
Memory
- Lorebook
- Alaya layer implementation
- Dreaming agent (memory management, long term memory soliation, personality update)
- Muscle memory (hard-coded skill library)
- Time awareness (scheduling & planning)
- Fix Apache Arrow's (upstream) support for MONTH_DAY_NANO intervals @sumimakito
- Adopt by DuckDB WASM
-
Body
- VRM (3D) (big shout out to @Lilia-Chen, love it!)
- VRM model positioning (1a2a790, feat(stage-ui): VRM model postioning functionalities #302)
- VRM model uploading functionality (d2ce04b, feat(stage-ui): Live2D vs VRM page switch botton and VRM model uploading functionality #290)
- VRM vs Live2D switching (d2ce04b, feat(stage-ui): Live2D vs VRM page switch botton and VRM model uploading functionality #290)
- Live2D (2D)
- Model scaling (fffe3d1, feat(stage): change scale and position of live2d model #275)
- Model positioning (fffe3d1, feat(stage): change scale and position of live2d model #275)
- Talking head @sumimakito @nekomeowww (read also: https://github.com/warmshao/FasterLivePortrait)
- Motion embedding & RAG from semantic (@nekomeowww @luoling8192 @sumimakito)
- Working in
moeru-ai/motion-gen(private repo, invite open) - Generating
- Labelling
- Storing / Retrieving
- DuckDB WASM /
pglitequery embedded vectors through ANN/KNN - Query function
- DuckDB WASM /
- Morphing / Lerp
- Live2D
- VRM
- Working in
- VRM
- LipSync
- VRM (3D) (big shout out to @Lilia-Chen, love it!)
-
Protocols
- MCP connecter
- Server SDK to configure child modules
- Server Event to configure child modules
- Web / Tamagotchi client configure child modules
-
Inputs
- Telegram video messages
- Discord text messages
- Perhaps design an IM layer? (Perhaps take a look at Koishi)
- Default included Speaches (a.k.a.
fast-whisper-server) for STT (from @kwaa) - bilibili danmaku
- YouTube live chat
- Twitch live
- Whiteboard (from @LemonNekoGH on comment)
-
Intelligence
- Multi-modal
- Speech
- Transcription
- 🤗 Transformers.js provider
- Speech (open for everyone to join)
- Chat (open for everyone to join)
- Scene manipulation
- Pre-defined object models callout (place in scene)
- Realtime 3D object models generation
- Light effects manipulation
- Stage images / backgrounds manipulation
- MCP launcher
- Three.js MCP
- Pixi.js MCP
- Browser use
- browse YouTube
- [Trial] Sing a song (@qyl27)
- Dataset preparation
- Trial & experiment
- Multi-modal
-
Games
- More games? What games wanted?
- Improve Minecraft with newly added DuckDB WASM memory layer @luoling8192 @nekomeowww
- Minecraft with CV capabilities
- Minecraft mod support (make it playable, try stream with it)
- Improve Factorio (make it playable, try stream with it)
- Factorio mod support
- Connect to Nintendo Switch (read also: https://zenn.dev/noov/articles/8afaf41678dfa7)
-
Engineering (big shout out to @kwaa)
- Turborepo integration (perf: use turborepo for build #212)
- Buildless architecture for UI packages (perf(ui): buildless #215, perf(ui-transitions): buildless #217, perf(ui-loading-screens): buildless #216, perf(stage-ui): buildless #218, perf(packages/ccc): buildless #211)
- Rolldown-vite adoption (032a44c, perf(apps): use
rolldown-viteinstead ofvite#214) - CI/CD Pipeline
- Nightly build support (fd31405, feat(ci): support nightly build #256)
- Build-only mode (5987b3c, feat(ci): build only #257)
- Artifact management improvements (16df338, a57c71d)
- CUDA Toolkit included
- Import sorting enforcement (ac9ec95)
-
unocss-preset-fontsand fonts improvements (7e0035c, fix(workspace): missing fonts #269, 43dc9a1, thanks to @Weathercold) - @moeru/std integration (c13811b, refactor: use
@moeru/std#226) - AVIF image format migration (76e5463, chore: use AVIF instead of PNG/JPG as more as possible #301)
-
Characters
-
Community Support / Marketing
- Stream with Airi (DevStream)
- DevLog 2025.07.18 publication (fcb7b14, docs(devlog): DevLog @ 2025.07.18 #289)
- Chronicles content creation (a878e9c)
- Internationalization & i18n
- Now i18n in
packages/i18n(eb6bcae, feat(i18n): new package for locales #249) - English
- Chinese (zh-Hans) documentation (cb0d767)
- Contribution guide i18n (05e83d6, i18n(docs): contribution guide #285, thanks to @Lilia-Chen)
- Now i18n in