research-summarizer

---
name: research-summarizer
description: Use when summarizing a supplied paper, article, or report into a structured brief, comparing multiple sources, or extracting and formatting citations.
category: Product
version: 1.0.0
tools: []
---

# Research Summarizer

Turn dense source material into actionable briefs: extract what matters, compare across sources, and format citations correctly. Read less, understand more, cite correctly.

## Scope

This workflow summarizes documents the user already has — papers, articles, reports pasted or attached. It does not search the web and needs no external service. It is distinct from: academic literature *discovery* (finding papers via search), entity due-diligence built from live web research, and open-ended "research [topic]" requests that require searching for sources. If the user wants sources *found* rather than digested, that's a different workflow — use web_search directly instead.

## When This Applies

"Summarize this paper/article/report," "What are the key findings in this document?," "Compare these sources," "Extract citations from this," "Give me a research brief on [topic]" (when a source is attached), "Break down this whitepaper" — any request to summarize, brief, review, cite, or compare supplied source material.

## Workflow 1 — Single-Source Summary

1. **Identify source type** and pick the matching structure: academic paper → IMRAD (Introduction, Methods, Results, Analysis, Discussion); web article → claim-evidence-implication; technical report → executive-summary structure; documentation → reference-summary structure.
2. **Fill the brief** with every section populated directly from the source:

```
Title / Author(s) / Date / Source Type

## Key Thesis — 1-2 sentences: the central argument or finding
## Key Findings — numbered, each with its supporting evidence
## Methodology — how they arrived at these findings: data sources, sample size, approach
## Limitations — what the source doesn't cover or gets wrong
## Actionable Takeaways — what to do with this information
## Notable Quotes — direct quotes with page/section reference
```

3. **Assess quality**: source credibility (peer-reviewed vs. reputable outlet vs. primary vs. secondary), evidence strength (data-backed vs. anecdotal vs. theoretical), recency (still relevant), and bias indicators (funding source, author affiliation, methodology gaps).

## Workflow 2 — Multi-Source Comparison

1. Collect 2–5 sources.
2. Summarize each with Workflow 1.
3. Build a comparison matrix — one row per dimension (central thesis, methodology, key finding, sample/scope, credibility rating), one column per source; no source skipped.
4. Synthesize: where do sources agree (convergent findings = stronger signal)? Where do they disagree (divergent findings = needs investigation)? What gaps exist across all of them? What does the weight of evidence favor?
5. Produce a synthesis brief: Consensus Findings, Contested Points (with the strongest evidence on each side), Gaps, and a Recommendation based on the weight of evidence.

## Workflow 3 — Citation Extraction

1. Detect citations in the text: DOI, URL, author-year, and numbered-reference patterns, deduplicated.
2. Format the list in the requested style (APA 7 by default), catching anything a pattern-based pass would miss.
3. Classify each: primary source (original research/data), secondary source (reviews, meta-analyses, commentary), or tertiary source (textbooks, encyclopedias).
4. Output a sorted bibliography with classification tags.

Supported citation styles: APA 7 (default, social sciences/business), IEEE (engineering/CS), Chicago (humanities/history), Harvard (general academic), MLA 9 (arts/humanities).

## Quality Assessment Framework

Rate every source on four dimensions:

| Dimension | High | Medium | Low |
|-----------|------|--------|-----|
| Credibility | Peer-reviewed, established author | Reputable outlet, known author | Blog, unknown author, no review |
| Evidence | Large sample, rigorous method | Moderate data, sound approach | Anecdotal, no data, opinion |
| Recency | Published within 2 years | 2–5 years old | 5+ years, may be outdated |
| Objectivity | No conflicts, balanced view | Minor affiliations disclosed | Funded by interested party, one-sided |

Overall rating: 4 Highs = strong source, cite with confidence. 2+ Mediums = adequate, cite with caveats. 2+ Lows = weak, verify independently before citing.

## Proactive Triggers

Flag these without being asked: a source with no date (undated sources lose credibility points); a source that contradicts another source (surface the contradiction explicitly, never paper over it); a paywalled source (note limited access, suggest alternatives if known); only one source supplied for a requested comparison (ask for at least one more — comparison needs 2+); incomplete citation fields (flag missing year/author/title, never invent metadata); a source 5+ years old in a fast-moving field (warn about potential obsolescence).

## Verification Before Delivery

Before handing over any brief, check: every Key Finding cites a location in the source (section, page, or quote) — no unanchored claims; the extracted citation count matches the bibliography count in the output (investigate any gap); every source carries the four-dimension quality rating, with weak sources flagged rather than silently included; comparison matrices have one row per dimension and one column per source with none skipped; nothing was invented — missing metadata is marked "not stated," never filled in.

## Related Work

Quantitative analysis is a separate, complementary workflow — use this one for qualitative sources and the other for metrics. Competitive-landscape research is likewise complementary: this workflow handles individual source analysis, a competitive teardown handles the market landscape. This workflow typically feeds content production — summarize sources first, then write. Desk research here pairs with primary user research done elsewhere.