Run literature reviews over a folder of papers
Researchers point Claude Code at 50+ papers to extract bibliographies, find core cited works, and write relevance lists against a theme as files.
Done withClaude Code
- What they did
- The author sets up custom slash commands in Claude Code, and the research theme goes in CLAUDE.md. /lit-map extracts text from every PDF in a papers/ folder. It pulls each paper's objective, method, conclusion and keywords, then writes a High/Medium/Low relevance list to literature_map.md. /citation-network extracts the bibliographies, finds works cited across several papers, and ranks them.
- How it went
- The author claims /lit-map cuts 50 papers from 2-3 days to 30 minutes, and /citation-network cuts 10 papers from 4-5 hours to 15 minutes. These are before/after claims, not measured results.
- Worth knowing
- Scanned or image-only PDFs lower accuracy. The researcher still has to check the AI's scores and summaries, and should review the terms and policies before sending unpublished or copyrighted text to cloud services.
Try it yourself with Claude Code
Read every paper in the folder [path to my papers folder] and extract each one's bibliography into a file called bibliographies.md. Then identify the works cited most often across the papers, and write relevance_list.md ranking all papers by how relevant they are to [my research theme], with one sentence of reasoning each. Don't modify or delete any of the original files. You're done when both files exist and you've told me the top 10 core works.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: note.com · Undated
Related use cases
Turn tickets into merged pull requests
Pulls tasks from GitHub Issues or Linear, then runs Claude Code or Codex in isolated Kubernetes pods to carry each one through to a PR.
Click through macOS app builds to catch visual bugs
A developer used Claude Code computer use to compile, launch, and click through macOS app screens to find layout bugs.
Run an entire business from the terminal with Claude Code
Poster built a system that runs their whole business through Claude Code instead of Claude.ai; the specific tasks aren't detailed in the snippet.
Run persistent financial research with an agent
LangAlpha is a Claude Code-style agent for investing research that keeps persistent workspaces and memory so analysis such as models and comps can be updated across sessions.
Custom agent
Turn voice notes into structured Obsidian notes
Voice notes sent over Telegram become Obsidian thought notes with transcript, cleaned argument, linked context and a draft; Hermes also tracks recipes.
Generate updated research reports on followed companies
The poster built and open-sourced a custom stock-research agent that produces updated reports on companies they follow.
Custom agent
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.