wtf is image-to-campaign?

image-to-campaign is a skill for Claude Code that turns a single product photo into a complete marketing funnel. Give it one image and a short interview, and it produces a Brand Brief covering product identity, color palette, fonts, tone, and calls to action. You review and approve that brief before anything else happens, so no generation credits are spent until you say yes. Once approved, the skill builds three connected pieces: a scroll driven landing page, three vertical Instagram Reels with ready to paste captions, and a styled blog article. Because everything comes from the same approved brief, the landing page, Reels, and blog article share one visual and verbal identity instead of looking like three separate AI outputs. The landing page is built first, and its hero video gets reused, reframed to a vertical format, for one of the Reels, which saves on generation cost. The written copy defaults to German, and the call to action is a placeholder for a demo, not a real shop. Both defaults can be changed in the skill's configuration file if a different language or a real conversion goal is needed. To actually generate images and videos, you need a Higgsfield account with credits, plus the Higgsfield command line tool, which the skill installs automatically (on Windows this requires a Bash shell such as Git Bash). You also need a browser and a simple static file server to preview and check the landing page locally, and general internet access since the generated pages pull fonts and animation libraries from content delivery networks. Everything else, meaning the orchestration logic, the landing page and Reel specifications, and templates, ships inside the skill itself, so no extra project files are required. The project also recommends installing companion Higgsfield skills for better asset quality and model selection. Output lands in a single campaign folder containing the approved brief, the website files, the three Reels with captions, and the blog files, ready to use or hand off.