What problem does it solve?
The novelty skill helps you avoid duplicating existing research by cross-verifying a proposed idea against web results, Semantic Scholar, and your internal wiki, then producing a conservative novelty score with actionable differentiation guidance.
Core Features & Use Cases
- Multi-source novelty verification: cross-checks your idea using WebSearch, Semantic Scholar, DeepXiv (hybrid search), arXiv recent preprints, and internal wiki papers/concepts/ideas.
- Review LLM cross-verification: when not in quick mode, an independent review model assesses novelty based on the collected prior work.
- Optional persistence to your wiki: with --write for an idea slug, saves only the resulting novelty_score to the idea page frontmatter and logs the write.
Quick Start
Run the novelty check for a wiki idea slug like sparse-lora-for-edge-devices and get a 1–5 novelty score with closest prior work recommendations.