literature-review
Skill3 starsChanged 42 days agoMITView on GitHub
Produces a thematic literature review that organizes a field into an argument - consensus, disputes, evidence quality, and the open gap the reader should care about. Use when someone asks "write my related-work section", "review the literature on X", "what does the research say about X", or "organize these papers into themes". Do NOT use for an exhaustive, protocol-registered review with formal inclusion criteria and risk-of-bias appraisal - use systematic-review instead. Do NOT use for synthesizing interviews or mixed non-academic sources into decision themes - use research-synthesis instead.
90 other projects use a version of this.
Adopted by
Most recently taken up by
- Ivan-propio/claude-skillstheir own version, first seen Nov 16, 2026
- tatashbar/Economic-Modeling-Studiotheir own version, first seen Nov 16, 2026
- lengrufeng2008/dsh-skill-librarytheir own version, first seen Nov 15, 2026
- PerryLink/dsh-research-skillstheir own version, first seen Nov 15, 2026
- ranveerankhira-gif/research-agent-skills-zhtheir own version, first seen Nov 15, 2026
- xiong-zx/skillstheir own version, first seen Nov 15, 2026
Used with
Projects with it also use pdf-explore, paper-narrative and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
JimLiuSkillpdf-explore
Use this skill when the user has attached a PDF, paper, report, or other document and the answer needs content from more than one place in it: summarize the methods or any other section, compare sections, find where a topic is discussed, read a value or label off a figure or chart, or find/list/extract every instance of something across the whole document (datasets, benchmarks, citations, figures, table rows, accession numbers — including appendices). Skip it only for a single lookup of 1–4 pages quoted in your very next response — `read_file(pages=[...])` attaches pages as images that are dropped from context after one turn, so multi-section answers end up re-reading the same ranges repeatedly. Parses the PDF once in the Python kernel: `pdf_pages` (pages as persistent text), `pdf_outline` (TOC), `pdf_scan` (rank pages by relevance), `pdf_map`/`pdf_extract` (per-page summary / structured fields via parallel haiku calls). For PDF creation/manipulation, use reportlab/pypdf directly.
Also in 9 projects that use this
JimLiuSkillpaper-narrative
Judge and reshape the STORY a paper's figures tell. Input is the work itself — manuscript (or abstract) + figure deck — no hand-written brief. `derive_paper_brief(abstract, captions)` extracts pitch/vision/per-figure-claims; a handling-editor reviewer on the full deck returns hook_verdict (would Fig 1 make me send this for review?), arc (hook→mechanism→evidence→application), figure_moves (panels in the wrong figure), missing_panels (concrete analyses to RUN), kill_list, and boldest_defensible_fig1. Hands per-figure claims to `figure-composer`. Load when writing or revising a paper.
Also in 5 projects that use this
Norman-burySkillbrainstorming-research
Use before writing any paper - clarifies paper type, discipline, topic, methodology, and chapter structure through natural dialogue
Also in 4 projects that use this
lingzhi227Skilllatex-formatting
Handle LaTeX formatting, templates, and styling for academic papers. Set up conference templates (ICML, ICLR, NeurIPS, AAAI, ACL), fix formatting issues, manage packages, and ensure venue-specific compliance. Use when the user needs to set up a paper template, fix LaTeX formatting, or prepare for submission.
Also in 4 projects that use this
lingzhi227Skillnovelty-assessment
Assess research idea novelty through systematic literature search. Multi-round search-evaluate loops with harsh critic persona. Binary novel/not-novel decision with justification. Use before committing to a research direction.
Also in 4 projects that use this
- +6 more from JimLiu
- +1 more from lingzhi227
Other versions90 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 90
- getcompanion-ai/feynman★ 10ksame name
- advaitpaliwal/feynman★ 10ksame name
- companion-inc/feynman★ 10ksame name
- Norman-bury/research-writing-skill★ 3.4ksame name
- Norman-bury/articlewriting-skill★ 3.4ksame name
- HughYau/AcademicForge★ 2.6ksame name
- K-Dense-AI/claude-scientific-writer★ 2.6ksame name
- xuzhougeng/wisp-science★ 1.1ksame name
- wu-yc/LabClaw★ 1.1ksame name
- LeonChaoX/qinyan-academic-skills★ 910same name
80 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository5 other files it gives its agents.
- academic-essaySkill
- citation-trackerSkill
- fact-checkerSkill
- policy-briefSkill
- systematic-reviewSkill
The file
We split this skill into the rules it gives an agent
8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 70 edited versions we read keep these.
8
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 57
Every cited claim must trace to a real, checkable source.
- Distinctive: line 65
Do NOT summarize one paper after another - that is an annotated bibliography, and it hides the disagreements that make a review useful.
- Distinctive: line 66
Do NOT paper over contradictory evidence to tell a tidy story.
- Distinctive: line 67
Do NOT treat citation count or repetition as truth - popularity is not evidence quality.
- Distinctive: line 68
Do NOT report only conclusions.
- Distinctive: line 69
Do NOT claim completeness ("all studies show") - a thematic review is representative, not exhaustive.
- Distinctive: line 73
No section is organized author-by-author.
- Distinctive: line 76
Every citation is real and traceable.
“Common” is measured: kept by at least 6 in 10 of the 70 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Literature Review
A literature review is an argument about the state of knowledge, not a list of summaries. The failure it prevents is the annotated bibliography in disguise - one paragraph per paper, no comparison, no position - which readers skip and reviewers reject because it demonstrates reading, not understanding. Every section must compare works against each other and end somewhere: a consensus, a live dispute, or a gap.
Operating procedure
Step 1: Gather inputs
Collect before writing:
- The guiding question or claim the review serves (a thesis chapter, a paper's related-work section, a research proposal, a decision memo). Default framing if none given: "What is known, what is contested, and what is missing about X?"
- The corpus: papers supplied, or search scope (fields, venues, timeframe) if collection is part of the job.
- Audience and length target (a related-work section runs 800-1,500 words; a standalone review 3,000+).
- Inclusion posture: representative coverage is the default; if the user says "every study, defensibly complete," stop and route to systematic-review - that job needs a protocol.
Label assumed scope as a guess and confirm.
Step 2: Scope and collect
Define the question, rough inclusion criteria, and timeframe. Prioritize two classes of work: seminal papers (highly cited, field-defining - the ones later work argues with) and recent work (the current frontier). A review citing only old classics is stale; only recent preprints, rootless. Snowball: mine reference lists of the 3-5 most central papers backward, and their citing papers forward.
Step 3: Extract claims, not summaries
For each work record: the central claim, the method, sample or setting, the key limitation, and which conversation it joins. One or two sentences per field. This table is what makes Step 4 possible - grouping requires comparable atoms.
Step 4: Organize by theme, never by author
Group works by approach, finding, or dispute - not chronologically and not paper-by-paper. A theme is a sentence, not a topic word. Bad theme: "Remote work studies." Good theme: "Remote work raises measured output but weakens team cohesion, with the effect moderated by tenure." Aim for 3-6 themes; more than that means the question is too broad or the themes are actually sub-points.
Step 5: Synthesize within each theme
For every theme, state three things explicitly:
- Consensus: what most credible work agrees on.
- Disputes: where findings conflict, naming both sides and hypothesizing why (different populations, definitions, methods, incentives).
- Evidence quality: weigh by rigor, not volume - three small convenience-sample studies do not outweigh one large well-designed one, and ten papers repeating one original finding is an echo, not ten pieces of evidence. Note methodology and limitations, not just conclusions.
Contrast pair for the synthesis sentence:
Bad: "Smith (2019) found X. Jones (2021) found Y. Lee (2022) also studied this topic."
Good: "Field experiments consistently find X (Smith; Lee), but survey-based studies find the opposite (Jones) - plausibly because self-report captures intention rather than behavior, a gap no study has yet tested directly."
Step 6: Find the gap
The gap is the point of the review. State explicitly: questions no work answers, populations or contexts not studied, methods not yet applied, and claims everyone repeats on thin evidence. If the review supports the user's own research, the gap must lead directly to their question - that alignment is the whole rhetorical job of a related-work section.
Step 7: Verify citations
Every cited claim must trace to a real, checkable source. Never fabricate or guess citations; flag anything unverifiable. For formatting and DOI verification of the final reference list, hand off to citation-tracker.
Deliverable
Produce a review containing: thematic sections that compare and contrast works, an explicit consensus/dispute/quality assessment per theme, a synthesis that takes a position on where the field stands, an explicit "gaps and future directions" section, and accurate citations throughout.
Do NOT
- Do NOT summarize one paper after another - that is an annotated bibliography, and it hides the disagreements that make a review useful.
- Do NOT paper over contradictory evidence to tell a tidy story; surfacing conflict is the review's highest-value output.
- Do NOT treat citation count or repetition as truth - popularity is not evidence quality.
- Do NOT report only conclusions; a finding without its method and limitations cannot be weighed.
- Do NOT claim completeness ("all studies show") - a thematic review is representative, not exhaustive; exhaustive claims require systematic-review discipline.
Quality bar
- No section is organized author-by-author; every paragraph compares at least two works.
- Each theme states consensus, disputes, and evidence quality - not just findings.
- The gap section exists and names specific unanswered questions, not "more research is needed."
- Every citation is real and traceable; none invented or approximated.
- A reader finishing the review could state the field's position and its biggest open question in two sentences.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

