# Contributing Thanks for helping make General Visual Content Studio more useful, predictable, and easier to adopt. ## Good contributions - Add a reusable visual route for a real class of briefs. - Improve plain-language brief interpretation without hard-coding one subject. - Add regression coverage for a routing, validation, or documentation failure. - Contribute a complete, rights-cleared example content pack. - Clarify installation, tool routing, or quality-control guidance. Please open a feature request before starting a large behavioral or structural change. ## Local setup Requirements: - Python 3.10 or newer - Pillow ```bash git clone https://github.com/Wang-Yeah623/Visual-studio-skill.git cd Visual-studio-skill python -m pip install pillow python -m unittest discover -s tests -v python skill/culture-visual-content-studio/scripts/validate_project.py \ --project examples/fragrance-kv-content-pack \ --strict ``` ## Skill-change checklist When changing `skill/culture-visual-content-studio/SKILL.md` or its references: 1. Start from a representative plain-language brief. 2. Record the expected route and the failure being prevented. 3. Keep subject, visual route, and mood as separate decisions. 4. Prefer general rules over brand- or object-specific prompt recipes. 5. Add or update a regression test where behavior can be checked mechanically. 6. Update the README and changelog when the public workflow changes. ## Example and asset policy Every contributed example should be reproducible and safe to redistribute. - Use original, permissively licensed, public-domain, or clearly generated assets. - Record source URL, author, license, retrieval date, local path, and SHA-256 where applicable. - Never submit confidential client assets or files scraped without permission. - Do not imply endorsement by a referenced brand. - Label generated or conceptual campaign work clearly. ## Pull requests Keep pull requests focused. Complete the pull request checklist and explain: - the brief or failure that motivated the change; - the expected user-visible behavior; - the commands and visual evidence used to validate it; - any compatibility or licensing considerations. By contributing, you agree that your contribution is licensed under this repository's [MIT License](LICENSE).