Snowline uses the Godot Engine, available under the MIT license.
Godot Engine: https://godotengine.org/
License and copyright notices: https://godotengine.org/license/
Godot's third-party license list: https://godotengine.org/license/#thirdparty

All game-specific models, UI graphics and audio in this project are generated
by the source code included in this project. No PaperRoute assets or code are used.

BAOSHI GEOGRAPHIC DATA
Map data © OpenStreetMap contributors, ODbL 1.0.
https://www.openstreetmap.org/copyright
https://opendatacommons.org/licenses/odbl/1-0/
Terrain Tiles / USGS SRTM and GMTED2010.
https://registry.opendata.aws/terrain-tiles/
See assets/baoshi/ATTRIBUTION.txt for data provenance and approximation notes.

REFERENCE CHARACTER EQUIPMENT
This independent Wuhuarou snowboard edition uses an unbranded white helmet,
raised blue goggles, camouflage outfit, lavender boots and one teal snowboard,
procedurally approximated from the project owner's reference video.
No POC or HEAD sponsor credit is shown in this edition. The reference video
and its soundtrack are not distributed with this game.

CHINESE FONT
Noto Sans SC Bold: https://github.com/notofonts/noto-cjk
Copyright Adobe / Noto contributors; SIL Open Font License 1.1.
Complete license: assets/fonts/OFL.txt

PHOTOGRAPHER VOICE
The short Chinese encouragement was generated locally using macOS speech
synthesis. No user photos or video were uploaded for speech generation.

Snowline UI Bold is a renamed subset derived from Noto Sans SC Bold,
licensed under SIL Open Font License 1.1. Copyright notices and full license
are retained in assets/fonts/OFL.txt and the font metadata.

BLUEBIRD DAY MUSIC
Original procedural composition generated by tools/generate_bluebird_music.py.
All tones and percussion synthesized locally; no third-party samples or recordings.

Keketuohai theme song (keketuohai_theme.wav): AI-generated using Volcengine Doubao Seed Audio 1.0 from project-authored lyrics and production instructions, at the project owner's request. Instrumental style is a synthetic approximation, not a traditional performance recording. Third generated candidate, trimmed at 109.9 seconds with a short fade to omit terminal prompt tones; downmixed to mono/24kHz for mobile playback. No reference singer recording was sampled or voice-cloned. Original AI-generated source retained locally.