Agent-Skills.md

Agent Skills: Literature Review

Conduct comprehensive, systematic literature reviews using multiple academic databases (PubMed, arXiv, bioRxiv, Semantic Scholar, etc.). This skill should be used when conducting systematic literature reviews, meta-analyses, research synthesis, or comprehensive literature searches across biomedical, scientific, and technical domains. Creates professionally formatted markdown documents and PDFs with verified citations in multiple citation styles (APA, Nature, Vancouver, etc.).

UncategorizedID: drshailesh88/integrated_content_OS/literature-review
2

Install this agent skill to your local

pnpm dlx add-skill https://github.com/drshailesh88/integrated_content_OS/tree/HEAD/skills/cardiology/literature-review

Skill Files

Browse the full folder contents for literature-review.

Download Skill

Loading file tree…

skills/cardiology/literature-review/SKILL.md

Literature Review

Comprehensive, systematic literature reviews following rigorous academic methodology across biomedical and scientific domains.

Triggers

  • User asks for a literature review on a topic
  • User needs evidence synthesis across multiple studies
  • User wants to understand the current state of research
  • User is preparing background for a newsletter or editorial
  • User needs to identify research gaps

Workflow Phases

1. Planning & Scoping

Define Research Question using PICO framework:

  • Population: Who is being studied?
  • Intervention: What treatment/exposure?
  • Comparison: Against what?
  • Outcome: What results matter?

Establish Scope:

  • Date range (typically 5-10 years for currency)
  • Study types to include (RCTs, observational, reviews)
  • Inclusion/exclusion criteria
  • Languages accepted

2. Systematic Search

Required Databases (minimum 3): | Database | Best For | |----------|----------| | PubMed | Biomedical, clinical | | Cochrane Library | Systematic reviews | | EMBASE | Pharmacology, European | | Google Scholar | Gray literature, coverage | | bioRxiv/medRxiv | Preprints, cutting edge |

Search Strategy Elements:

  • MeSH/controlled vocabulary terms
  • Free-text synonyms
  • Boolean operators
  • Field restrictions as needed

3. Screening & Selection

  1. Deduplication - Remove duplicate records
  2. Title screening - Exclude obviously irrelevant
  3. Abstract review - Apply inclusion criteria
  4. Full-text evaluation - Final eligibility check
  5. Document exclusion reasons at each stage

4. Data Extraction

Extract from each included study:

  • Citation information
  • Study design and setting
  • Sample size and characteristics
  • Key findings and effect sizes
  • Funding sources
  • Limitations noted by authors

5. Quality Assessment

| Study Type | Assessment Tool | |------------|-----------------| | RCTs | Cochrane Risk of Bias (RoB 2) | | Observational | Newcastle-Ottawa Scale | | Systematic reviews | AMSTAR 2 | | Diagnostic | QUADAS-2 |

6. Synthesis

Critical: Organize thematically, NOT study-by-study.

Bad: "Smith 2020 found X. Jones 2021 found Y. Brown 2022 found Z."

Good: "The relationship between SGLT2 inhibitors and heart failure outcomes has been consistently demonstrated across multiple trials (Smith 2020, Jones 2021, Brown 2022), with effect sizes ranging from..."

Identify:

  • Areas of consensus
  • Contradictory findings and potential explanations
  • Research gaps
  • Methodological trends

7. Citation Verification

Before finalizing:

  • Validate all DOIs against CrossRef
  • Verify author names and dates
  • Confirm journal names
  • Use citation-management skill for formatting

Output Formats

For Newsletter Background

  • Concise synthesis (500-1000 words)
  • Key statistics highlighted
  • Top 5-10 most relevant citations
  • Clear clinical implications

For Editorial Foundation

  • Comprehensive synthesis (1500-2500 words)
  • Organized by theme/controversy
  • Evidence quality assessment
  • Explicit gaps identification

For Academic Manuscript

  • Full systematic review format
  • PRISMA flow diagram
  • Risk of bias tables
  • Forest plots if meta-analysis

Best Practices

  1. Document everything - Search strategies, screening decisions, extraction forms
  2. Search multiple databases - Minimum 3 for comprehensive coverage
  3. Include preprints - For most current findings
  4. Assess study quality - Don't treat all evidence equally
  5. Synthesize thematically - Never list studies sequentially
  6. Acknowledge limitations - Search constraints, publication bias
  7. Verify every citation - Accuracy is non-negotiable