SkillFoxx

literature-survey-skill

A Claude Code skill for literature surveys: from a pile of papers to a view of the field via NotebookLM and /survey modes.

  • 115 ★ on the repository
  • Last commit
  • MIT
Skill

Extended access

Why this level

  • The skill sends papers to NotebookLM through MCP and works with your Google account.
All reasons and checks

snl-ucsb/literature-survey-skillAuthor:SNL-UCSB

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/literature-survey-skill

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The skill sends papers to NotebookLM through MCP and works with your Google account.

This is third-party code. Review the repository files before installing.

What it does

The skill hands off the plumbing of a literature survey: finding papers, extracting claims, cross-referencing terms and building comparison tables. NotebookLM ingests the papers and returns summaries with exact quotes, while Claude Code runs the workflow, fills templates and tracks reading depth. The researcher forms hypotheses and decides what matters. Five modes run as commands: /survey intent captures goals and survey type, triage maps the landscape and prioritizes reading, deepen reads a paper in three passes with claim extraction, evidence audit and writing-craft analysis, synthesize compares papers for the final document and expand proposes growing the corpus. The skill draws on three-pass reading and Kahneman-style bias checks.

Who it is for. PhD students and researchers running a survey of dozens of papers who want a structured process instead of chaotic reading.

Good fit when

  • A new field has a hundred papers and needs a map of directions with reading priorities.
  • One key paper must be read deeply: claims, evidence, weak points.
  • Approaches from different papers must be compared on the same dimensions before writing a section.
  • You must check that an idea that seems new is not already covered in the corpus.

Not a fit when

  • You need a quick summary of one paper without process or templates.
  • Materials cannot be uploaded to a Google service because of publisher or lab restrictions.
  • Your field is far from systems and networking and you are not ready to adapt the skill.

Example request

Run /survey intent: I am surveying network protocols for a paper, with 40 PDFs and two months.

Limitations

The skill needs a configured NotebookLM MCP server and a Google account sign-in; details are in NOTEBOOKLM_SETUP.md. Papers are uploaded to NotebookLM, so consider the terms on their distribution. Defaults are tuned for systems and networking venues, and for other fields templates and references need editing. The skill fits surveys of 10 to 150 papers, and on small samples the process is excessive. The researcher still verifies the results.

How to disable. Delete the ~/.claude/skills/survey folder. Disconnect the NotebookLM MCP server and the Google sign-in separately in Claude Code settings.

FAQ

How do I install the skill?

Clone the repository and run `./setup.sh`, which copies the skill to `~/.claude/skills/survey/`. Manually, copy `SKILL.md`, `reference/` and `templates/`.

What are the Explorer, Investigator, Validator and Examiner survey types?

Four archetypes with different defaults, from a landscape overview of 50-100 papers to a full exam survey of 80-150 papers.

Is NotebookLM required?

Yes, the skill relies on the NotebookLM MCP CLI for grounded summaries. It is registered as an MCP server in Claude Code.

What does deepen mode do?

It takes a paper through two or three passes: claims with quotes, an evidence audit, a decomposition by state, time, coordination and interface, then writing-craft analysis.

Which venues is the skill tuned for?

Defaults are tuned for systems and networking, such as SIGCOMM, NSDI, CoNEXT and IMC. The Customize for Your Field section explains adapting it to another field.

Discussion

No questions or reviews yet.

Skills

Five skills for literature reviews: question refinement, modules, method comparison, contradictions and the research gap.

SkillMinimal access89

A Claude Code skill for research papers: a five-stage pipeline, section rules, a style audit and an independent red-team review.

SkillMinimal access249

35 skills for Claude Code and Codex lead a social scientist from idea to a journal manuscript, with scans of participant data.

PluginBroad access190

A Claude Code skill for scientific literature reviews: an interview, Scopus and OpenAlex search, DOI checks and a final docx file

SkillExtended access371

Claude Code skills for the academic pipeline: research, paper drafting, peer review and citation checks

SkillExtended access56.7K

A skill for writing and revising ML, CV and NLP papers: section structure, paragraph flow and reviewer-style self-review

SkillMinimal access8.3K

README in short

The repository consists of the skill, reference files on reading methods and templates for notes, intent, triage and synthesis. Modes are described with examples of what each step produces. The skill is written first of all for systems and networking researchers.

README badge

Original SKILL.md

---
name: survey
description: "Literature survey assistant with four modes: /survey intent (capture student goals, expertise, and success criteria), /survey triage (landscape mapping via NotebookLM), /survey deepen (structured reading with craft and visualization extraction), /survey synthesize (cross-paper analysis for related work, area exams, gap identification). Also: /survey expand (corpus growth proposals). Use when students mention reading papers, literature review, related work, area exam prep, paper corpus, or any systematic reading task."
---

# Literature Survey Skill — Intent → Triage → Deepen → Synthesize

This skill helps PhD students build a deep, synthesized corpus of research insights through a structured, cognitively-aware workflow. It combines Kahneman's dual-process theory, Keshav's three-pass reading method, first-principles analysis, and NotebookLM as a backend query engine.

**Architecture:** All student work lives locally (Obsidian/filesystem). NotebookLM is a query backend only — papers go in, grounded answers come out.

**Modes:**
- `/survey intent` — Capture what you're trying to learn and where you're starting from
- `/survey triage` — Map the landscape, prioritize reading depth
- `/survey deepen` — Structured reading with craft and visualization extraction
- `/survey synthesize` — Cross-paper analysis for deliverables
- `/survey expand` — Structured corpus growth proposals

If the student invokes `/survey` without a mode, ask which mode they want. If they seem unsure or are starting a new survey, begin with intent.

**Prerequisites:** NotebookLM MCP CLI must be configured. See `reference/notebooklm_tools.md` for setup.

---

## Mode 1: Intent — "What are you trying to learn?"

**Purpose:** Capture the student's survey goals, expertise level, and success criteria BEFORE any paper is read. This shapes all subsequent modes.

**Do NOT create any NotebookLM notebooks in this mode. Do NOT ingest any papers. The goal is purely clarifying intent.**

### Step 1: Identify the survey archetype

Ask the student which best describes their situation:

> Before we look at any papers, I need to understand what kind of survey this is. Which best describes you?
>
> **A. Explorer** — "I'm entering a new area and need to understand the landscape."
> **B. Investigator** — "I have specific questions I need answered from the literature."
> **C. Validator** — "I think I've found a gap/idea and want to confirm it's novel."
> **D. Examiner** — "I need to demonstrate comprehensive mastery for an exam or survey paper."

Each archetype has different defaults:

| Archetype | Triage scope | Deepen targets | Pass 3 count | Synthesize output |
|-----------|-------------|----------------|-------------|-------------------|

Report a problem