What problem does it solve?
This skill removes the manual burden of discovering, screening, and synthesizing academic literature by orchestrating a reproducible two-pass review workflow that combines wide search, agent-driven screening, fetch planning, and structured synthesis into a persistent session.
Core Features & Use Cases
- Two-pass workflow orchestration: init → search → screen → fetch → agent read & synthesize, with persistent session state stored per review.
- Source integration and deduplication: queries OpenAlex (no API key) and optionally arXiv, dedups by arXiv ID, DOI, and normalized title.
- Fetch planning for full-text reading: suggests arxiv-analyze, direct PDF URLs, or marks abstract-only items so the agent can prioritize reads.
- Use case: run a scoped search for a research question, have the agent screen 50 candidates to a 10–25 paper shortlist, then produce a structured literature review with TL;DR, established findings, debates, and gaps.
Quick Start
Start a literature review session by initializing with a clear research question and then run search to gather candidates.
