Claude Study Skills banner
adityakumar68 adityakumar68

Claude Study Skills

Development community

Description

Claude Code skills for actually learning things — Socratic tutoring, lecture breakdowns, depth-calibrated explanations, and critical paper reviews

Installation

This entry records only its repository, not the path inside it, so there is no exact command to give. Open the source below and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

README

claude-study-skills

Claude Code skills for actually *learning* things — built by a PhD student tired of tutorial hell.

The default way AI assistants explain things — the complete, polished answer, all at once — feels like learning but doesn't survive a retrieval test a week later. These skills flip that: you do the cognitive work (recall, explain, predict, derive), Claude steers, diagnoses, and corrects.

The skills

Skill What it does
tutor Adaptive Socratic tutoring sessions. Assesses what you already know first, teaches one idea at a time (every turn ends in a question you must answer), makes you explain the topic back, corrects misconceptions explicitly, and only declares mastery after real checks — explain-back, a transfer question, and (where it applies) reproducing the derivation or code from memory. Tracks everything across sessions with spaced reviews (+1/+3/+7/+21 days).
lecture-map Breaks a lecture, course, YouTube video, slide deck, or textbook chapter into a dependency-ordered checklist of studyable topics — prerequisites / core / teased-but-not-taught — each with a depth target (intuition, derivation, or implementation) and a handoff into a tutor session. Reads real PDFs/PPTX/DOCX, supports page ranges ("map pages 30–58"), and anchors each topic to its source slides/pages.
explain-mode Explanations at a deliberately chosen depth — ELI5 through with-the-maths — with jargon defined at first use and a diagnosis of which step broke when a previous explanation didn't land.
paper-breakdown Critical breakdowns of research papers: what problem it solves, whether the claims hold up, whether the baselines were fair, and a read/skim/skip recommendation — not a restatement of the abstract.

`tutor` and `lecture-map` share a `study/` workspace (an existing `study/` in your working directory or a parent, el