Bloom skill
A personal AI tutor for Claude Code: a syllabus, one lesson at a time, and each next lesson shaped by your annotations and feedback
- 294 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skill creates course folders, Markdown files and a learning-log.jsonl journal in the working directory
- The web version sends lesson text and uploaded files to the LLM provider, and the API key lives in .env
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/bloomDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill creates course folders, Markdown files and a learning-log.jsonl journal in the working directory
- The web version sends lesson text and uploaded files to the LLM provider, and the API key lives in .env
This is third-party code. Review the repository files before installing.
What it does
Bloom turns Claude Code into a tutor for any topic. You say what you want to learn and the bloom-tutor skill creates a course folder with a syllabus and a first lesson in Markdown. You read the lesson, mark confusing passages with ???, write feedback at the end and say you are done. The skill writes the next lesson from your answers, your marks and the gaps found in the previous one. Syllabus depth comes in three levels: simple, standard and deep. When every syllabus item is checked off, the skill gives a final assessment and then a summary. The plugin carries six more learn-* skills, from judging whether a topic is worth the effort to checking understanding by Feynman-style retelling. Next to it sits a React and FastAPI web version. It builds courses from a topic, an uploaded file or project files.
Who it is for. For people learning a new field in Claude Code who want to study by a syllabus with lessons in order.
Good fit when
- You need to work through an unfamiliar topic step by step: a syllabus first, then one lesson at a time
- You want to leave ??? marks right in the lesson text and get answers in the next lesson
- You want to check whether a topic is worth the effort before studying it: learn-occam does that
- You have a PDF or notes and want to ask questions on highlighted passages, which the web version supports
Not a fit when
- You need lessons in English or Russian: bloom-tutor replies in Chinese only
- You want a quick answer to one question: by its rules the skill answers with a question of its own and leads you to the conclusion
- You want every lesson at once: the skill creates one document at a time and waits for your feedback
Example request
Create a new folder and help me learn the basics of graph theoryLimitations
The bloom-tutor skill works in Chinese whatever the request language. The Claude Code mode needs nothing else. The web version needs Python 3.11 or newer with uv, Node.js 18 or newer and a key for any OpenAI-compatible API, such as DashScope or OpenAI. You can bring it up with `docker compose up -d`, and the interface opens on port 3000. A course lives only in Markdown files in the working folder, and the skill has no separate database. The repository is MIT-licensed.
How to disable. Remove the plugin through /plugin or delete skills/bloom-tutor from .claude/skills. Courses are ordinary folders of Markdown files; delete them separately.
FAQ
How do I install Bloom in Claude Code?
Run `/plugin marketplace add Li-Evan/Bloom` and `/plugin install bloom@li-evan`. Alternatively clone the repository and copy skills/bloom-tutor into .claude/skills.
What are the ??? marks in lessons for?
A mark on a confusing passage gives the skill a snapshot of your thinking. It reads every ??? and the feedback at the end before writing the next lesson.
How does a course end?
Once every syllabus item is checked off, the skill runs a final assessment and only then writes the summary. You cannot ask for the summary earlier.
How is the web version different from the skill?
In the web version you read lessons in a browser and ask questions about highlighted text, with a study calendar and next-topic suggestions. A course can be built from an uploaded PDF, TXT or MD.
Which skills are in the Bloom plugin?
Seven. bloom-tutor runs a whole course, learn-deep runs five lenses on a topic at once and learn-crossover explains the new through the familiar. learn-occam judges the return on learning, and learn-graph builds a knowledge map. learn-prototype teaches through a crude prototype, and learn-feynman checks understanding by having you explain it back.
Discussion
No questions or reviews yet.
Related
SkillsA Claude Code skill: name a topic and it researches the web and builds a course of two-minute audio lessons
A mentor skill for Claude Code that guides you to LeetCode solutions with hints instead of handing over code at once
A Claude Code plugin that catches your corrections in chat and, after review, writes them into CLAUDE.md and AGENTS.md
Four free Claude Code skills: an IELTS coach that grades writing, breaks down reading mistakes and prepares the speaking part
Claude Skills (Alireza Rezvani)
Claude Code Skills & Plugins
A large library of skills, agents and commands by domain: engineering, marketing, product, compliance, C-level advisors and research
An exam prep skill built on your own lectures and problems: the agent goes through the course chapter by chapter with citations
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| Bloom (this entry) | no | 294 | |
| Engram | no | 1,466 | a learning engine with first-principles curricula and free recall |
| Learning Opportunities | no | 2,529 | offers short exercises during ordinary coding work |
| AI Engineering from Scratch | no | 75,591 | a ready-made AI engineering course with tutor skills, the topic is fixed |
README in short
The README opens with Benjamin Bloom's 1984 study: one-on-one tutoring lifts results by two standard deviations. A quick start for the CLI and web modes follows, then a table of skills.
README badge
Original SKILL.md
--- name: bloom-tutor description: Use when 用户想以一对一苏格拉底导师的方式系统学习一个课题——开一门新课、推进课题的下一篇、提交学习反馈或说「我读完了」、或整理/查看学习日志。基于 Bloom 2 Sigma 的交互式学习系统。触发词：开个文件夹学X、我想学X、帮我学X、继续、下一篇、我读完了、整理学习、查看学习日志、interactive Socratic tutoring、Bloom 2 sigma learning。 --- # Bloom Tutor · 交互式苏格拉底学习系统 ## 这是什么 基于 Benjamin Bloom「2 Sigma Problem」研究（1984）的一对一 AI 导师系统。每个课题是一个独立文件夹，通过**自适应生成的课程文档 + 用户反馈循环**模拟一对一苏格拉底式导师，把学习效果推向 +2σ。学习的主要载体是文档，对话只是辅助确认状态。 ## 永远中文 所有回复、解释、提问、文档一律使用中文。 ## 工作守则（不可违背） 触发本 skill 后，以下守则在整个学习交互全程生效——**违反字面就是违反精神**： 1. **每次只生成一篇文档。** 输出后必须等用户读完并反馈，才能生成下一篇。无论用户怎么要求，绝不一次性批量生成多篇（如 `01.md`+`02.md`+`03.md`）。 2. **启动新课题必须在同一轮内**生成 `syllabus.md` + 首篇 `01.md`，不拆成两轮，不先做任何苏格拉底诊断提问——用户会在 `01.md` 反馈区给出理解情况，你据此再调整。 3. **大纲学习深度**支持「简单 / 标准 / 深入」三档；用户未指定时默认「标准」，具体条目范围见 `references/syllabus.md`。 4. **用户不能主动触发 `summary.md`。** 任何「总结一下」「生成总结」类请求，统一回应：「总结会在你学完所有掌握项后自动生成，现在还没到时候。」 5. **生成任何新文档前必读**：该课题所有已有 `.md` + 文末「你的反馈」+ 全文所有 `???`/`？？？` 标注。