Agent.index

Research & Knowledge

Run a systematic literature review from one prompt

SciSpace's agent generates PICO questions, searches and screens papers, downloads PDFs, extracts data and outputs the full review with a PRISMA diagram.

Done with ChatGPT Agent

What they did

User enters the prompt at scispace.com. The agent generates PICO-based research questions and screening criteria for the user to review and edit, then lets the user select databases and date ranges. After that the agent searches papers across databases, reranks results, performs title/abstract and full-text screening, downloads PDFs, extracts data, and produces the review plus PRISMA diagram and inclusion table.

How it went

It outputs a complete systematic review and PRISMA diagram from a single prompt, but paywalled papers still block full automation.

Worth knowing

The main catch is papers behind paywalls.

Try it yourself with ChatGPT Agent

Run a systematic literature review on [research question]. Write PICO questions, search [databases, e.g. PubMed, Semantic Scholar], screen papers against inclusion criteria [criteria], and extract [data points] into a table. Finish with a written review and a PRISMA-style flow summary of how many papers were found, screened and included, and ask me to confirm the inclusion criteria before screening.

Source on X - Apr 12, 2026

Read the original