11 skills · category

Research & Analysis skills for Claude Code

Skills for careful research: structured and systematic literature reviews, critical reading of papers, citation management, market sizing with sources, survey design, qualitative coding, reproducible computational research, fact-checking with primary sources, and technology evaluations that end in a decision record.

Download all 11 (.zip) Browse the list

Every research & analysis skill

Each card opens the full playbook, example prompts and install commands.

No. 481

Academic Writing

Papers where every claim matches its evidence

Claude drafts papers in the order that keeps the argument honest: contribution sentence first, then figures, methods and results, and the introduction and abstract last. It turns related work from a list into a comparison, reports statistics with effect sizes and confidence intervals, matches hedging to evidence, verifies every citation, and drafts point-by-point reviewer responses and venue-specific AI-use disclosures.

academic-writing 238 lines · 2 files

No. 482

Citation Management

Clean bibliographies built from DOIs, not memory

Claude audits your .bib with a bundled script that flags duplicate keys, preprint-plus-published duplicates, missing fields and malformed DOIs, then rebuilds entries from doi.org and Crossref instead of guessing. It cross-checks cited against defined keys, sets up Zotero with Better BibTeX auto-export, and renders any CSL or BibLaTeX style with zero undefined citations.

citation-management 176 lines · 2 files

★ No. 483

Fact-Checking

Trace every claim to a primary source you opened

Claude splits a text into atomic claims, reads laterally to judge each source, and follows every citation trail back to the dataset, paper or transcript it came from. Each claim gets a verdict from accurate to fabricated with a confidence label and a re-checkable link, and AI-generated text gets extra scrutiny for invented citations, quotes and numbers.

fact-checking 162 lines

No. 484

Literature Review

Reproducible searches, verified sources, argued gaps

Claude frames the question with PICO or a concept-context split, builds Boolean strings tested against known gold papers, logs every database run, screens against criteria written in advance, and chases citations to saturation. Findings are synthesized by theme from an extraction matrix, gaps are tied to empty cells, and every reference is checked against doi.org or Crossref before it ships.

literature-review 197 lines · 2 files

No. 485

Market Research

TAM you can defend, every number with a source

Claude defines the market and the decision first, sizes it bottom-up from buyer counts and real price points, cross-checks with a top-down estimate, and explains any gap instead of averaging it away. Every figure carries a source tier, year and confidence label, assumptions sit in a table you can change, and vendor press-release CAGRs never drive the base case.

market-research 162 lines · 2 files

No. 486

Paper Analysis

Map every claim to its evidence before judging

Claude reads a paper in three passes, builds a claim map that ties every headline claim to the table or figure behind it, and rates the evidence strength. It checks the design against a checklist suited to the study type (trial, observational, ML benchmark, survey, qualitative), works through threats to validity, and ends with a calibrated verdict instead of a paraphrased abstract.

paper-analysis 160 lines

No. 487

Qualitative Coding

Codebooks, kappa checks and themes backed by quotes

Claude picks reflexive or codebook thematic analysis to match the study, builds codes with definitions, include and exclude rules and counter-examples, and checks double-coded samples with a bundled kappa script that names the exact disagreeing segments. Themes are written as claims, tested against negative cases, and every excerpt is verbatim with a source ID.

qualitative-coding 163 lines · 2 files

No. 488

Reproducible Research

One command from raw data to every figure

Claude sets up a research repo where raw data is immutable and versioned with DVC, environments are locked with uv, conda-lock or renv, and a declared pipeline regenerates every figure and manuscript number. It pins seeds and tolerances, keeps notebooks honest, proves the result with a clean-room rerun, and archives code and data with DOIs.

reproducible-research 204 lines

No. 489

Survey Design

Questions that map to decisions, samples that hold up

Claude maps every question to a decision before writing it, rewrites double-barreled, leading and vague items, picks fully labeled scales, and orders the flow so data quality holds. A bundled calculator sizes the sample from the smallest subgroup you need, with finite-population correction and design effect, and the analysis plan and exclusion rules are written before fielding.

survey-design 167 lines · 2 files

No. 490

Systematic Reviews

Protocol first, PRISMA counts that actually add up

Claude writes a registrable protocol before any screening, builds PRESS-checked search strategies, runs dual independent screening with kappa checks, and extracts data with page-level locators. A bundled script computes PRISMA 2020 flow counts from your screening log and flags missing reasons or impossible decisions, so the diagram, bias tables and GRADE ratings stay consistent.

systematic-reviews 178 lines · 3 files

No. 491

Technology Evaluation

Pick tools with evidence, a PoC and an exit plan

Claude fixes requirements and weights before naming candidates, eliminates options with pass/fail gates, and scores the shortlist with evidence pulled from docs, release history and repo health rather than memory. A time-boxed PoC targets the riskiest assumption, TCO and exit cost are priced, a sensitivity check exposes close calls, and the result lands as an ADR.

technology-evaluation 163 lines

Choosing Research & Analysis skills

Install the skills that match requests you make every week; each costs about 100 tokens of standing context until it loads. If you are not sure where to begin, start with Fact-Checking. For background, read The Best Claude Code Skills in 2026, by Use Case, How to Install Claude Code Skills (5 Ways) and Personal vs Project Skills.

Install several

Install several at once

The installer accepts any number of skill names. Add --project to install into the current repository.

curl -fsSL https://claudeskills.page/install.sh | sh -s -- academic-writing citation-management fact-checking literature-review market-research paper-analysis qualitative-coding reproducible-research survey-design systematic-reviews technology-evaluation

Research & Analysis skills FAQ

How do I install every Research & Analysis skill at once?

Download the bundle at https://claudeskills.page/downloads/category/research.zip and unzip it into ~/.claude/skills/ (or .claude/skills/ inside a repository). To pick a few, pass several names to the installer: curl -fsSL https://claudeskills.page/install.sh | sh -s -- academic-writing citation-management fact-checking.

Will installing 11 skills slow Claude Code down?

No. Claude Code keeps only each skill's short description in context — roughly 100 tokens per skill — and loads the full instructions only when a request matches. Install the whole category if it fits your work; unused skills cost almost nothing.

Which Research & Analysis skill should I start with?

Start with Fact-Checking. They cover the most common requests in this category; add the rest as your work calls for them.