Recommendation Engine Skill for Cold-Start Product Similarity
Use Claude Code to build an explainable content recommendation baseline from product tags when interaction history is sparse.
中文Install in Claude Code
Use a synthetic catalog, normalized tags, and a Python script that computes set overlap.
- Get the complete `plugins/ai-ml/recommendation-engine/skills/building-recommendation-systems` directory from the pinned GitHub commit.
- Copy it into the target project's `.claude/skills/`. Check `.claude/skills/building-recommendation-systems/SKILL.md`, start Claude Code at the project root, and call `/building-recommendation-systems`.
Example task
Query a new mug tagged coffee and ceramic. Candidate A adds travel, B has coffee and glass, and C has tea and glass. Exclude the query item and rank by Jaccard similarity.What to check
The order is A, B, C with scores 2/3, 1/3, and 0. The output lists the overlapping tags behind the first result.
Original SKILL.md on GitHub · Updated
Frequently asked question
Can I recommend items without browsing history?
Build a content similarity baseline from item attributes, then compare collaborative or hybrid methods once interaction data is available.