Research Summarizer

What this skill helps you do

Turns dense source material—academic papers, web articles, technical reports, and documentation—into structured, actionable briefs. Supports three workflows: single-source summaries (IMRAD for papers, claim-evidence-implication for articles), multi-source comparisons with synthesis matrices identifying consensus and contested points, and citation extraction with five output formats (APA, IEEE, Chicago, Harvard, MLA). Includes a quality assessment framework rating every source on credibility, evidence strength, recency, and objectivity. Built for product managers, analysts, and founders who read more than they should have to.

Install and get started

Copy the full instructions into your AI tool. Test one low-risk example before connecting real store data.

Original Skill instructions

You are a Research Summarizer. Process provided documents—no web search needed. Three workflows: (1) Single-source: identify type (academic→IMRAD, article→claim-evidence, report→executive). Use format_summary.py --template <type> to scaffold, then fill Key Thesis, Findings with evidence, Methodology, Limitations, Takeaways, Quotes. (2) Multi-source: summarize each, build comparison matrix, synthesize agreements/disagreements, produce consensus findings with weight-of-evidence. (3) Citation extraction: run extract_citations.py, format in APA/IEEE/Chicago/Harvard/MLA. Rate every source on Credibility, Evidence, Recency, Objectivity. Never invent metadata—mark missing as 'not stated'. Flag sources >5yr old in fast-moving fields.

Useful tasks

• Summarizing academic papers into structured executive briefs for product teams
• Comparing multiple research sources to find consensus and contested findings
• Extracting and formatting citations from documents in APA, IEEE, or Chicago style
• Assessing source quality across credibility, evidence, recency, and objectivity
• Creating research briefs from whitepapers, technical reports, and industry analyses

How to use it

• Use format_summary.py to scaffold templates—fill from source, don't write from scratch
• Rate every source on all four quality dimensions; flag weak sources explicitly
• For comparisons, build a matrix with one row per dimension, one column per source
• Run extract_citations.py and verify its total matches your bibliography count
• Mark missing metadata as 'not stated' rather than inventing details

More skills for this workflow

• Shopify Support Triage Agent

  Instantly categorize and prioritize customer tickets

• Returns & Refund Processing Assistant

  Automate the entire returns workflow from request to resolution

• Customer Success Manager

  SaaS health scoring, churn prediction, and expansion detection via Python CLI tools.

• Ecommerce Customer Service Pro

  Industry-selectable ecommerce CS: pre-sale, after-sale, complaints, and KOL communication.