litreview
Skill1 starsChanged 5 months agoMITView on GitHub
Run a structured academic literature review — scope the question, search the literature, screen for relevance, extract findings into a comparable matrix, synthesize themes and gaps, and cite cleanly (BibTeX). Use for research papers, related-work sections, grant background, or surveying a field before building. Tracks what was searched so the review is reproducible, not cherry-picked.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- ericluo04/claude-academic-workflowtheir own version, first seen Oct 9, 2026
- Hassaan146/claude-skillstheir own version, first seen Jul 25, 2026
- mahiunfiltered/claude-code-skillstheir own version, first seen Jul 8, 2026
- Hussain-Elfalah/Agentic-Skillstheir own version, first seen Jun 4, 2026
- ByteBaker/claude-skillstheir own version, first seen Mar 20, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- ericluo04/claude-academic-workflow★ 25same name
- Hassaan146/claude-skills★ 18same name
- Hussain-Elfalah/Agentic-Skills★ 2same name
- mahiunfiltered/claude-code-skills★ 1same name
- ByteBaker/claude-skills★ 0same name
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 repository24 other files it gives its agents.
- api-designSkill
- brainstormingSkill
- councilSkill
- dispatching-parallel-agentsSkill
- document-toolkitSkill
- executing-plansSkill
- gui-input-automationSkill
- humanizeSkill
- hyprland-controlSkill
- incremental-implementationSkill
- issue-to-prSkill
- mcp-builderSkill
- rust-reviewSkill
- screen-captureSkill
- session-handoffSkill
- shadcn-auditSkill
- spec-to-issuesSkill
- subagent-driven-developmentSkill
- teachSkill
- unslopSkill
- using-git-worktreesSkill
- visual-computer-use-loopSkill
- writing-flowSkill
- writing-plansSkill
The file
We split this skill into the rules it gives an agent
3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.
0
Common
Most of the 5 edited versions we read keep these.
2
Distinctive
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 30
Keep a search log: query | source | date | hits | kept.
- Distinctive: line 65
No fabricated citations: Every reference must resolve to a real paper — verify the DOI/arXiv id.
- Distinctive: line 70
If coverage is partial (paywalls, time), say so explicitly.
“Common” is measured: kept by at least 6 in 10 of the 5 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 review is only as good as its method. Make the search reproducible and the synthesis honest about what's missing — don't just collect papers that agree.
1. Scope the question
Pin down before searching: the precise question, inclusion/exclusion criteria (years, venues, methods, language), and what "done" looks like (N papers? a saturation point where new searches surface nothing new?). Write these down — they make the review defensible.
2. Search
Cast wide across sources; record each query and source so it's reproducible:
- Semantic Scholar (
api.semanticscholar.org) — has a free API for title/abstract/citation search and citation graphs. - arXiv (
export.arxiv.org/api) — preprints, esp. CS/ML/physics. - Google Scholar / OpenAlex / PubMed — broader / domain-specific.
- Citation chasing — follow references (backward) and citations (forward) of the key papers. Often finds more than keyword search.
Keep a search log: query | source | date | hits | kept.
3. Screen
Two passes: title/abstract first (drop the clearly irrelevant), then full-text on survivors against your criteria. Note why each rejection happened — that's part of the method.
4. Extract into a matrix
One row per paper, comparable columns. Save as a table (Markdown or CSV):
| Cite key | Year | Problem | Method | Data | Key finding | Limitations | Relevance |
Extract claims faithfully; quote exact numbers; note the paper's own stated limitations. If you can't get the full text, mark it — don't infer findings from the abstract alone and present them as established.
5. Synthesize
Organize by theme or approach, not paper-by-paper (a list of summaries isn't a review). For each theme: what's the consensus, where do papers disagree, what's the trend over time. End with gaps — what's unanswered, under-studied, or methodologically weak. The gap is usually the point, especially for a paper's related-work or a grant's background.
6. Cite
Collect BibTeX as you go (Semantic Scholar / arXiv / DOI give it directly). Keep
a references.bib. Cite every claim that isn't yours. Match the target venue's
style.
Honesty rules
- No fabricated citations. Every reference must resolve to a real paper — verify the DOI/arXiv id. Never invent authors, titles, or numbers.
- Distinguish what a paper showed from what it speculated.
- Report the search, including what you didn't find — a gap you didn't search for isn't a gap you can claim.
- If coverage is partial (paywalls, time), say so explicitly.
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.