Use this skill when the user wants a systematic literature review, survey, or synthesis across multiple academic papers on a topic. Also covers annotated bibliographies and cross-paper comparisons. Searches arXiv and outputs reports in APA, IEEE, or BibTeX format. Not for single-paper tasks — use academic-paper-review for reviewing one paper.
Systematic Literature Review is a research-synthesis skill that produces a structured survey across many academic papers on a topic. It solves the problem of manually gathering and comparing large bodies of literature: given a topic query, it searches arXiv, extracts structured metadata (research question, methodology, key findings, limitations) from each paper in parallel, synthesizes themes across the whole set, and emits a final report with consistent citations in APA, IEEE, or BibTeX format. It is deliberately scoped for breadth-first synthesis and routes single-paper deep reviews to a separate academic-paper-review skill.
The workflow has five phases: plan (confirm topic, scope up to a hard cap of 50 papers, citation format, and output location), search (call a bundled Python arXiv script that handles URL encoding, Atom XML parsing, and id normalization, with strong guidance to use 2-3 core keywords and category filters rather than long queries), extract metadata in parallel via delegated subagents, synthesize, and write the report. The skill is careful about tool discipline: it mandates using the bundled read-only search script rather than writing an ad-hoc HTTP client, instructs running the search exactly once to avoid wasted calls, forbids fabricating paper metadata on failure, and requires reporting the true result count rather than padding the list. It ships templates for each citation format and an evals set.
Target users are researchers, graduate students, and technical writers who need a literature survey, annotated bibliography, or cross-paper comparison. Use cases include surveying a subfield, comparing methodologies across recent papers, and producing a citation-formatted synthesis over a defined time window.
Use it for breadth-first synthesis across many papers: literature surveys, cross-paper comparisons, or annotated bibliographies. For deep review of exactly one paper it routes you to the separate academic-paper-review skill.
It searches arXiv via a bundled Python script. The default scope is 20 papers with a hard upper bound of 50; beyond that it advises splitting the survey by sub-topic because synthesis quality degrades.
APA, IEEE, and BibTeX, each with a bundled template. APA is the default when the user does not specify a venue.
It reports the actual count rather than padding the list, runs the search only once instead of retrying variations, and if the script fails it tells you the error and stops rather than fabricating paper metadata.
No. It requires using the bundled arxiv_search.py script, which handles URL encoding, Atom XML parsing, and id normalization, rather than scraping arXiv or writing a custom HTTP client.
Quick Setup:
.claude/skills/Repository
bytedance/deer-flow