SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsliterature-review

literature-review

Writing & research

View repository

Updated 10-16 06:50

About this skill

Literature Review Writing

Synthesize multiple academic papers into a cohesive literature review with thematic organization, trend identification, gap analysis, and proper citation weaving.

When to Use

  • Writing the Introduction or Background section of a manuscript or grant
  • Creating a standalone literature review paper
  • Synthesizing findings from an opencite search session
  • Identifying gaps in the literature to justify new research
  • Preparing a systematic or scoping review

Review Types

Type Purpose Method Output
Narrative Summarize and interpret a body of literature Selective, thematic Flowing prose with arguments
Systematic Exhaustive, reproducible search with defined criteria PRISMA protocol, inclusion/exclusion Structured report with methodology
Scoping Map the extent of literature on a topic Broad search, categorization Overview of themes and gaps
Mini-review Focused review for a specific question Targeted search Brief, focused synthesis

Review Pipeline

Step 1: Gather Papers

Use the opencite skill to find relevant literature:

# Foundational papers
uvx opencite canonical "topic" --max 15

# Recent work
uvx opencite search "specific query" --max 30 --sort date

# Citation network
uvx opencite cite "key-paper-DOI" --direction both

# Download and convert for reading
uvx opencite batch-fetch --from-json results.json --convert -o ./papers --summary report.json

Step 2: Read and Annotate

For each paper, extract:

  • Key finding (1-2 sentences)
  • Method (approach/technique used)
  • Population/context (who/what was studied)
  • Limitations noted by authors
  • How it connects to other papers in the set

Create a synthesis matrix:

Paper Finding Method Supports Contradicts
Smith 2024 X increases Y RCT, N=200 Jones 2023 Lee 2022
Jones 2023 X correlates with Y Survey, N=1000 Smith 2024 --
Lee 2022 No effect of X on Y Meta-analysis -- Smith 2024, Jones 2023

Step 3: Identify Themes

Group papers by theme, not chronologically. Common organizing structures:

  • By methodology: group studies using similar approaches
  • By finding: group studies with converging/diverging results
  • By population: group by subject type or context
  • By theoretical framework: group by underlying model
  • Chronological evolution: trace how understanding changed over time (use sparingly)

Step 4: Write the Review

Structure Per Section/Paragraph

Each paragraph in a literature review should:

  1. Topic sentence: State the theme or claim this paragraph addresses
  2. Evidence: Cite 2-5 papers supporting or nuancing this claim
  3. Synthesis: Compare, contrast, and connect the cited works
  4. Transition: Lead to the next theme or the gap

Citation Weaving Patterns

Integral citation (author as subject): "Smith et al. (2024) demonstrated that X increases Y using a randomized controlled trial with 200 participants."

Non-integral citation (finding as subject): "Recent evidence suggests that X increases Y in controlled settings (Smith et al., 2024; Jones et al., 2023), although this effect may be population-dependent (Lee et al., 2022)."

Synthesis citation (multiple sources for one claim): "Several studies have converged on the finding that X modulates Y through mechanism Z (Smith et al., 2024; Jones et al., 2023; Chen et al., 2021), with effect sizes ranging from d = 0.3 to d = 0.8."

Contrast citation: "While Smith et al. (2024) found significant effects using direct measurement, Lee et al. (2022) reported null results in a meta-analysis of self-report studies, suggesting that measurement approach may moderate the effect."

Avoid Common Pitfalls

  • Annotated bibliography: Don't just summarize each paper sequentially. Synthesize.
  • String-of-pearls: Don't write "Smith found X. Jones found Y. Lee found Z." Connect them.
  • Overclaiming: Don't say "proves" or "definitively shows." Use "suggests," "indicates," "provides evidence for."
  • Recency bias: Include foundational/seminal work, not just the last 2 years.
  • Confirmation bias: Include contradictory findings and explain discrepancies.

Step 5: Identify the Gap

After synthesizing existing literature, articulate what is missing:

Gap statement patterns:

  • "Despite these advances, no study has examined {specific question}."
  • "While X has been established for {context A}, it remains unclear whether this extends to {context B}."
  • "Existing studies are limited by {methodological issue}, and {new approach} could address this."
  • "The relationship between X and Y has been studied in isolation, but their interaction under {condition} is unexplored."

Step 6: Generate Bibliography

# Export BibTeX for all cited papers
for doi in $(cat cited_dois.txt); do
  uvx opencite lookup "$doi" -f bibtex --append-bib review_refs.bib
done

Output Format

# Literature Review: {Topic}

## {Theme 1 Title}

{Synthesized paragraphs with citations}

## {Theme 2 Title}

{Synthesized paragraphs with citations}

## {Theme 3 Title}

{Synthesized paragraphs with citations}

## Current Gaps and Future Directions

{Gap analysis and research opportunities}

## References

{BibTeX file or formatted reference list}

Quality Checklist

  • Organized thematically (not paper-by-paper)
  • Each paragraph synthesizes multiple sources
  • Citations woven into prose (not parenthetical dumps)
  • Contradictory findings acknowledged and explained
  • Seminal/foundational work included
  • Recent work (last 2-3 years) included
  • Gap statement clearly articulated
  • Transition sentences connect paragraphs and sections
  • No overclaiming (appropriate hedging)
  • All cited papers in reference list

Additional Resources