from medical-research-skills349
Fetch and summarize Scite.ai Smart Citations for a DOI to classify supporting, contrasting, and mentioning citations and assess a paper's reception.
Provides a packaged workflow to query Scite.ai Smart Citations for a given DOI and summarize how the paper is cited (supporting, contrasting, mentioning). It produces reproducible outputs useful for literature review, claim validation, and comparative analysis. The skill exposes CLI scripts to run batch or single DOI analyses and emits text or JSON outputs.
Use when evaluating a paper's reliability, prioritizing reading for literature reviews, fact-checking scientific claims, comparing citation sentiment across papers, or when a reproducible file-based result is required. It's intended for researchers and analysts working with DOIs and needing structured citation-classification counts.
scripts/scite_client.py (primary), plus several helper scripts for extraction and output formatting.Agents capable of running Python scripts and making external HTTP calls (requires Scite API availability).
This skill queries the Scite.ai API to classify citations for a given DOI as supporting, contrasting, or mentioning. The single script (scite_client.py) is a straightforward argparse CLI that calls a public endpoint. It exited with an arg-parsing error when run without a DOI argument (expected). SKILL.md is heavily duplicated with two overlapping 'When to Use' and 'Key Features' sections. No hardcoded secrets or destructive commands found, but the script spoofs a Chrome User-Agent string.
Legitimate research tool. Main issues are code quality (inconsistent naming, repetitive docs, User-Agent spoofing). No security red flags. Architecture is basic - single script, no references/ directory, duplicated SKILL.md sections.
