This is your bread and butter Pillow wrapper for when you need to resize, crop, convert formats, or slap a watermark on images without spinning up AI models. It's deterministic pixel pushing: make thumbnails, prep social media crops, batch process a directory, optimize for web delivery. The author clearly delineates when to use this versus their bria-ai skill, which is refreshing. If you're compositing product shots or doing semantic edits, you want AI. If you're just making that 4K image into three responsive variants and a WebP, this does it. The API is clean with sensible defaults, and it handles the usual sources like URLs, base64, and file paths without ceremony.
npx -y skills add bria-ai/bria-skill --skill image-utils --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
juliusbrussee/caveman
mattpocock/skills
shadcn/improve
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills