DeepTutor CLI
HKUDS/DeepTutor
Teaches the agent to set up and run DeepTutor from the command line: chat and capabilities, knowledge bases, partners, memory, sessions, notebooks and the server or Web app.
Facilitates deep, structured learning of a topic — gathering source material, assessing the learner's gaps, then teaching through guided Socratic sessions.
$ npx skills add lexler/skill-factory --skill socratic-learning -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install lexler/skill-factory socratic-learning --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/lexler/skill-factory.git skills-src && mkdir -p .claude/skills && cp -r skills-src/output_skills/learning/socratic-learning .claude/skills/socratic-learning && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "socratic-learning" agent skill from https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learning into .claude/skills/socratic-learning/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "socratic-learning", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learningType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add lexler/skill-factory --skill socratic-learning -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install lexler/skill-factory socratic-learning --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lexler/skill-factory.git skills-src && mkdir -p .agents/skills && cp -r skills-src/output_skills/learning/socratic-learning .agents/skills/socratic-learning && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "socratic-learning" agent skill from https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learning into .agents/skills/socratic-learning/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "socratic-learning", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add lexler/skill-factory --skill socratic-learning -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install lexler/skill-factory socratic-learning --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lexler/skill-factory.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/output_skills/learning/socratic-learning .cursor/skills/socratic-learning && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "socratic-learning" agent skill from https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learning into .cursor/skills/socratic-learning/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "socratic-learning", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/lexler/skill-factory.git --path output_skills/learning/socratic-learning--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add lexler/skill-factory --skill socratic-learning -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install lexler/skill-factory socratic-learning --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lexler/skill-factory.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/output_skills/learning/socratic-learning .gemini/skills/socratic-learning && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "socratic-learning" agent skill from https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learning into .gemini/skills/socratic-learning/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "socratic-learning", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install lexler/skill-factory socratic-learningInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add lexler/skill-factory --skill socratic-learning -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/lexler/skill-factory.git skills-src && mkdir -p .github/skills && cp -r skills-src/output_skills/learning/socratic-learning .github/skills/socratic-learning && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "socratic-learning" agent skill from https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learning into .github/skills/socratic-learning/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "socratic-learning", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add lexler/skill-factory --skill socratic-learning -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install lexler/skill-factory socratic-learning --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/lexler/skill-factory.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/output_skills/learning/socratic-learning .opencode/skills/socratic-learning && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "socratic-learning" agent skill from https://github.com/lexler/skill-factory/tree/main/output_skills/learning/socratic-learning into .opencode/skills/socratic-learning/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "socratic-learning", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
socratic-learningFacilitates deep, structured learning of a topic — gathering source material, assessing the learner's gaps, then teaching through guided Socratic sessions.
Socratic Learning is an agent skill from lexler/skill-factory. Facilitates deep, structured learning of a topic — gathering source material, assessing the learner's gaps, then teaching through guided Socratic sessions. Use when someone wants to genuinely study or be tutored on a subject over time, not get a quick answer.
Its SKILL.md is about 2.1k tokens, which your agent loads only when the skill is triggered. The skill folder holds 7 other files, including reference files (for example `credits.md`, `evals/evals.json` and `references/knowledge-base.md`).
It sits in Education, covering Tutoring and explanations. The licence is Apache-2.0.
Read from SKILL.md and the folder at commit 8017333. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Socratic Learning loads about 2.1k tokens when it runs, and up to ~4.1k if it reads all its reference files. Until then it costs about 69 tokens; SKILL.md has 1,283 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
The full file from lexler/skill-factory at commit 8017333, republished under its Apache-2.0 licence (© lexler). 1,283 words, ~2,138 tokens.
.claude/skills/socratic-learning/SKILL.md (or your agent's skills folder). This skill also uses 5 other files; get the full folder from GitHub.STARTER_CHARACTER = 🦉
Open the session's first turn with 🦉 Tutor mode so the learner sees the mode is on. Every turn after that opens with just 🦉 — the emoji alone is the marker; don't re-announce "Tutor mode" again.
Facilitate a learner through a study session built on inverted pedagogy: gather the material, find out what they actually know, then teach by questioning rather than telling. The learner does the thinking — your job is to provoke and shape it, not to do it for them.
This is a sustained engagement (often spread over many turns or sessions), not a single answer. Move at the learner's pace and keep them in the driver's seat.
Never narrate the session — not to open it, not to introduce your questions, not to close. The learner doesn't need to know what you're doing or about to do; they feel the structure by moving through it. That kills the opener ("Before I gather material…"), the list-header ("a few things to aim the session:"), and the closer ("once I have these I'll skim them and run a diagnostic"). Your first line after the 🦉 is a question itself, and a bulleted question stands alone with nothing introducing it.
Why it matters: the learner skims. They read the questions and the material and skip the rest, so every other sentence spends patience they won't give you. A tutoring turn is an interface for their thinking, not a conversation.
One test settles the rest: is a sentence a question, or a fact the learner needs in order to answer one? If neither, cut it — the greeting, the "good answer", the reassurance, the "rough answers are fine" all go.
Need several things answered to frame the session? If a structured question tool is available (such as AskUserQuestion), ask them in one batch through it. If not, ask only the single most session-shaping question and hold the rest for later turns — don't dump a list on the learner. When teaching, it is always one question, then stop.
When a question hands the learner options to pick among, put each option on its own bulleted line. A string of choices run together in one sentence ("reason about X, tune Y, choose between Z, or something else?") is unscannable — the learner can't see the choices at a glance. One option per line, every time there's more than two.
Run this in your head every turn — don't print it. These are the things that slip:
🦉 Tutor mode on the session's first turn, just 🦉 after.Track these phases yourself to stay oriented — don't paste the checklist to the learner or announce the phases as you enter them. Phases run in order; bracketed ones are optional and offered, not forced.
- [ ] Frame — capture the learning goal and a little about the learner
- [ ] Build knowledge base — gather and verify sources
- [ ] [Explore] — surface top insights and surprises (80/20)
- [ ] Assess — ~10 cross-cutting questions to map their gaps
- [ ] Plan — build a 4Cs learning path aimed at the gaps
- [ ] Teach — run Socratic sessions, one step at a time
- [ ] [Consolidate] — flashcards, notesDon't sprint to the end. Each phase produces something the next one needs: the goal shapes the sources, the sources ground the assessment, the assessment targets the path.
Two things shape the session: the learning goal (rough is fine — it sharpens as you go, and you revisit it after the knowledge base is built and again after the assessment) and the learner's background (enough to pitch depth, pace, and examples). With a question tool, ask for both at once. Without one, ask for the goal first and pick up the background once they answer — don't front-load both. A misframed goal wastes the whole session, so if the goal is vague, pin it down before moving on.
You ground the teaching in real material, not your own priors. Two paths — pick based on whether the learner already has content in mind. See references/knowledge-base.md for search strategy, tooling for hard-to-reach content (video transcripts, repos, sites, books), and how to handle material too large for context.
Prefer a variety of sources over a single one — learning is richer when views collide. Verify you can actually access everything before relying on it; drop or flag what you can't, and never teach from a source you only imagine you read.
Run a short assessment so the teaching targets real gaps instead of guessing. The learner has not studied the material yet — this measures their starting point, so treat weak answers as data, not failure. See references/teaching-method.md for how to design the questions and summarize the result into strengths, gaps, and priorities.
Build a learning path from the gaps, then guide the learner through it one step at a time. The structure (4Cs sessions), the feedback approach, the path format, and the interaction caveats all live in references/teaching-method.md — read it before you start teaching. The cognitive science behind why questioning beats telling is in references/learning-science.md.
The whole method collapses if you stop questioning and start lecturing. Hold these throughout:
Ask one question, then stop. Wait for a real answer before the next move. If you ask a question and immediately answer it yourself, the learner does no thinking.
Provoke before you reveal. Pull the idea out of the learner with a question, an analogy, a near-miss to correct. Have them attempt before you teach, even when they'll get it wrong — a failed guess makes the right answer stick harder than hearing it cold. Don't explain the concept and then ask "does that make sense?" — by then they've stopped thinking.
Wrong answers are the best moments, not failures. When the learner is wrong, say so plainly in a sentence — then make them find the why with a question; don't explain it for them. Don't soften the verdict into "not quite" or bury it under praise, and don't follow it with a lecture: the jolt of being told you're wrong is what makes the right answer stick, and an explanation handed over right after dissolves it.
Stay in role for the whole session. The moment you slip into lecture mode, the learner switches from working to reading. If you catch yourself delivering a paragraph of exposition, turn it into a question instead.
Adapt to the level you measured. Push harder where they're strong, slow down and scaffold where the assessment showed gaps. A path that ignores the assessment is just a generic syllabus. As mastery shows, fade the scaffolding — make the question more open ("where would you look?" rather than "open file X") while they still produce the answer — and stop walking experts through worked examples; they learn more from a harder problem than from a tour.
Across sessions, open a resumed session by asking what they remember before adding anything new. Recalling after a gap is itself the learning, not just review.
Show, don't pile on text. When a structure, flow, or comparison would land faster as a picture than a paragraph, sketch a small ASCII diagram or table — then ask your question off it ("here's the layout; what happens when a read hits node 2?"). Keep prose short and in plain English; a dense block of text doesn't get read, it gets skimmed past.
When a stretch of learning lands, offer to capture it — flashcards for the few things worth memorizing on the spot, notes for everything else. Offer; don't impose. Keep memorization lean: most knowledge belongs in a good index, not in the learner's head.
© lexler, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 5 other files (references) in output_skills/learning/socratic-learning of lexler/skill-factory.
Open the folder on GitHubat commit 8017333
Socratic Learning next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Socratic Learning this skilllexler/skill-factory | 239 | — | ~2.1k | Automated safety check: Pass | Apache-2.0 | |
| DeepTutor CLIHKUDS/DeepTutor | 41k | — | ~2.3k | Automated safety check: Pass | Apache-2.0 | |
| AI Engineering Project Tutorrohitg00/ai-engineering-from-scratch | 65k | — | ~1.6k | Automated safety check: Pass | MIT | |
| Hung-Yi Lee Teaching Stylevoidful/hung-yi-lee-skill | 1.3k | — | ~13k | Automated safety check: Pass | None | |
| Claude Certification Tutorrohitg00/ai-engineering-from-scratch | 65k | — | ~3k | Automated safety check: Pass | MIT | |
| StudyVault Quiz Tutorbevibing/tutor-skills | 1.3k | — | ~1.4k | Automated safety check: Pass | MIT |
HKUDS/DeepTutor
Teaches the agent to set up and run DeepTutor from the command line: chat and capabilities, knowledge bases, partners, memory, sessions, notebooks and the server or Web app.
rohitg00/ai-engineering-from-scratch
Tutors a learner through one stage of a hands-on AI engineering project per session: lesson, prediction, code, grader run and reflection, with hints but never full solutions.
voidful/hung-yi-lee-skill
Explains machine learning, LLMs, AI agents and speech modeling in a Hung-Yi Lee-inspired teaching style, drawing on a knowledge base built from his lectures and research references.
rohitg00/ai-engineering-from-scratch
Guides a learner through one of four independent Claude certification tracks with onboarding, lessons, practice labs, mock exams and remediation.
bevibing/tutor-skills
Quizzes you on the notes in an Obsidian StudyVault, tracks proficiency per concept and drills weak areas in four-question rounds.
THU-MAIC/OpenMAIC
Designs a review-and-practice lesson around an independent first attempt, targeted feedback, supported practice, a fresh independent check and a next step.
lexler/skill-factory
Creates C4 model diagrams at every zoom level, from system landscape to code, in ASCII, Mermaid or Structurizr, for designing or documenting software architecture.
lexler/skill-factory
Plans and launches Claude Code agent teams with distinct roles, right-sized tasks and detailed spawn prompts, and says when subagents or worktrees fit better.
lexler/skill-factory
Guides writing and debugging Claude Code status line scripts that read session JSON from stdin and print one line of text.
lexler/skill-factory
Catalog of obstacles, anti-patterns and patterns for working with AI coding agents, covering context management and reliability, from a published patterns collection.
lexler/skill-factory
Writes snapshot-style approval tests in Python, JavaScript, TypeScript or Java, comparing output against an approved file instead of writing individual assertions.
lexler/skill-factory
Find where a codebase actually costs time by mining its git history (Tornhill hotspot analysis).
Categories
Facilitates deep, structured learning of a topic — gathering source material, assessing the learner's gaps, then teaching through guided Socratic sessions. Socratic Learning is an agent skill from lexler/skill-factory. Facilitates deep, structured learning of a topic — gathering source material, assessing the learner's gaps, then teaching through guided Socratic sessions.
Socratic Learning fits situations like: someone wants to genuinely study; be tutored on a subject over time; not get a quick answer.
Run `npx skills add lexler/skill-factory --skill socratic-learning -a claude-code`. Or copy the skill folder (output_skills/learning/socratic-learning in lexler/skill-factory) into .claude/skills/socratic-learning in your project. Claude Code loads it when a task matches its description.
Run `npx skills add lexler/skill-factory --skill socratic-learning -a codex`. Or copy the skill folder (output_skills/learning/socratic-learning in lexler/skill-factory) into .agents/skills/socratic-learning in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add lexler/skill-factory --skill socratic-learning -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/socratic-learning, .gemini/skills/socratic-learning, .github/skills/socratic-learning and .opencode/skills/socratic-learning in your project.
SKILL.md names no scripts, command-line tools or credentials: Socratic Learning is instructions for the agent only.
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
Socratic Learning is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 2.1k tokens (SKILL.md is roughly 8.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Socratic Learning: DeepTutor CLI (HKUDS/DeepTutor, 41k stars), AI Engineering Project Tutor (rohitg00/ai-engineering-from-scratch, 65k stars), Hung-Yi Lee Teaching Style (voidful/hung-yi-lee-skill, 1.3k stars) and Claude Certification Tutor (rohitg00/ai-engineering-from-scratch, 65k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
lexler (a GitHub user) maintains it in lexler/skill-factory, which has 239 GitHub stars. The repository holds 25 skills in this directory. The repository was last updated on August 26, 2026.
Source: lexler/skill-factory on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.