Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
- Added instruction to prefix all user inputs with "please think very deeply" before processing. - No changes to the output format or core search workflow. - Documentation updated to describe this new input handling step.
What This Skill Does
Searches major academic databases to find and compile citation lists on a given topic. Covers Semantic Scholar, PubMed, arXiv, IEEE, ACM, and optionally Scopus and Web of Science. Returns deduplicated results with authors, title, venue, year, and DOI.
Searches across multiple indexes in one pass and deduplicates results, saving the back-and-forth of querying each database manually.
When to use it
- Building a Related Work section for a research paper
- Finding foundational papers before starting a literature review
- Collecting recent citations on a fast-moving topic like LLMs
- Gathering biomed references for a grant proposal
- Compiling a reading list on an unfamiliar research area
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Find recent papers on CRISPR off-target detection methods
- 1Asks user to confirm topic keywords, desired time range, and whether survey or recent papers are preferred
- 2Queries Semantic Scholar and PubMed using multiple synonym queries (CRISPR off-target, Cas9 specificity, guide RNA mismatch)
- 3Queries arXiv for preprints on the same terms
- 4Deduplicates results, preferring journal versions over preprints where both exist
- 5Returns a bullet list: Authors. Title. Venue. Year. DOI
A deduplicated citation list of 10-20 papers with consistent formatting, plus an offer to export as BibTeX or filter by subtopic
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.