literature-survey

Search academic indexes and generate a cited Markdown literature survey.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/UnaryLab/ai-for-research --skill literature-survey-unarylab
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: literature-survey
Source: https://github.com/UnaryLab/ai-for-research/tree/main/skills/literature-survey
Command: npx skills add https://github.com/UnaryLab/ai-for-research --skill literature-survey-unarylab

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires scripts/paper_db.py, scripts/search_semantic_scholar.py, scripts/search_arxiv.py, scripts/search_openalex.py, scripts/search_crossref.py, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Produces a rigorous, well-structured literature survey with verified, direct inline citations, so researchers can quickly understand the state of the art and write related work without citation errors.

Core Features & Use Cases

  • Multi-source paper search across Semantic Scholar, arXiv, OpenAlex, and Crossref, routed by venue family to improve coverage.
  • Disciplined 6-phase workflow (frontier → survey → deep dive → code/artifacts mapping → synthesis → final report), including mandatory full reading in deep dive.
  • Citation-safe survey writing that uses direct inline hyperlinks for every non-trivial claim and supports optional historical lineage tracing.

Quick Start

Run the skill on a topic like "AI accelerators for sparse attention" to generate a fully cited Markdown survey report with inline hyperlinks.

Frequently Asked Questions about literature-survey

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I write a literature survey with verified inline citations?▼

A literature survey with verified inline citations is generated by searching multiple academic indexes, running a phased deep-dive workflow, and synthesizing findings into a Markdown report with direct hyperlinks. This process ensures every non-trivial claim is backed by a linked source.

Can I search arXiv and Semantic Scholar to map the state of the art for a research topic?▼

Yes, you can map the state of the art by searching across arXiv, Semantic Scholar, OpenAlex, and Crossref. Searches are routed by venue family to improve coverage, synthesizing cross-family topics spanning AI/ML, computer architecture, and Nature-family science.

What's the best way to generate a related work section without citation errors?▼

The best way to generate a related work section without citation errors is using a citation-safe writing workflow that enforces direct inline hyperlinks. It requires a full reading during the deep dive phase to ensure citation integrity before synthesis.

How does a systematic review workflow handle historical lineage tracing?▼

A systematic review workflow handles historical lineage tracing by supporting optional historical lineage tracking during the synthesis phase. It follows a disciplined 6-phase process from frontier search to final report, mapping the evolution of research topics.

Does this literature survey tool work for cross-family topics outside of AI and ML?▼

Yes, the literature survey tool works for cross-family topics outside AI and ML. It is explicitly applicable to computer architecture and Nature-family science, routing searches by venue family to improve source coverage across different scientific domains.

What are the steps to synthesize academic papers into a cited Markdown report?▼

To synthesize academic papers into a cited Markdown report, follow a 6-phase workflow: frontier search, survey, mandatory full-reading deep dive, code and artifacts mapping, synthesis, and final report generation with inline-linked narrative and gated outputs.