# LearnOS

> LearnOS is a learning platform with a hosted MCP server. Learners study courses in a web reader with AI chat that cites the exact lesson section, spaced-repetition flashcards, notes, highlights, slide decks and audio overviews — and connect the same library to Claude, Cursor, or any MCP client to search and read it from their own AI tools.

## MCP server

- Endpoint: https://learnosapp.com/mcp (Streamable HTTP; MCP protocol 2025-11-25, 2025-06-18, 2025-03-26, 2024-11-05, 2024-10-07)
- Auth: OAuth 2.1 (browser consent); requests carry the user's session token, so Postgres row-level security applies as in the web app
- Discovery: https://learnosapp.com/mcp/server-card (MCP Server Card), https://learnosapp.com/.well-known/oauth-protected-resource
- Claude Code: `claude mcp add --transport http learnos https://learnosapp.com/mcp`
- Citations: search results name the course, lesson, heading, lesson_id and anchor, with a URL to that exact section
- Price: free; slide-deck and audio-overview generation count toward free usage limits

## Product

- [LearnOS MCP server](https://learnosapp.com/mcp-server): What the MCP server is, how to connect, tools, and FAQ
- [Discover](https://learnosapp.com/discover): Public course catalog
- [LearnOS for AI tools](https://learnosapp.com/ai): Using LearnOS courses from Claude, Cursor, Codex and Gemini
- [LearnOS for learners](https://learnosapp.com/students): Self-directed study with cited AI chat and flashcards
- [LearnOS for educators](https://learnosapp.com/educators): Publishing and assigning courses
- [Research](https://learnosapp.com/research): Published benchmarks, including ToolBench
- [Blog](https://learnosapp.com/blog): Articles and course write-ups

## Docs

- [Architecture](https://learnosapp.com/docs/architecture): How LearnOS works under the hood — one library, three surfaces, citation-grounded retrieval.
- [Discover](https://learnosapp.com/docs/discover): Browse shared courses and modules — stars, forks, saves, and model provenance.
- [Getting started](https://learnosapp.com/docs/getting-started): Sign up, add a course, and connect Claude or the CLI to your LearnOS workspace.
- [LearnOS](https://learnosapp.com/docs): An AI-native learning workspace with a hosted MCP server and CLI for driving it from Claude or your terminal.
- [Interactive blocks](https://learnosapp.com/docs/authoring/interactive-blocks): Reference for data-only fences that render as interactive lesson widgets
- [ask](https://learnosapp.com/docs/cli/ask): One-shot answer grounded in your course content, with citations.
- [Auth](https://learnosapp.com/docs/cli/auth): Sign in to the LearnOS CLI with browser OAuth or a personal access token.
- [CLI overview](https://learnosapp.com/docs/cli): Query courses, lessons, sections, and ask the knowledge base from your terminal.
- [lesson & chunk](https://learnosapp.com/docs/cli/lesson): Read full lessons and individual sections.
- [Presentations](https://learnosapp.com/docs/cli/presentations): Generate, share, and edit course slide decks from the CLI
- [Course quality](https://learnosapp.com/docs/cli/quality): Score a course's intrinsic content quality — citations, structure, content richness, retrieval-readiness, scaffolding, and writing.
- [search & courses](https://learnosapp.com/docs/cli/search): Discover courses and run hybrid search over lesson content.
- [Notes & flashcards](https://learnosapp.com/docs/cli/write): Add flashcards and notes to the lessons in your library.
- [Video lessons](https://learnosapp.com/docs/cli/youtube): Read the timestamped transcript behind a video lesson.
- [MCP overview](https://learnosapp.com/docs/mcp): A hosted MCP server exposing your LearnOS workspace to Claude and other MCP clients.
- [Install](https://learnosapp.com/docs/mcp/install): Connect Claude Code, Claude Desktop, or Cursor to your LearnOS workspace via the hosted MCP endpoint.
- [Prompts](https://learnosapp.com/docs/mcp/prompts): The learnos-ask and learnos-quiz-me slash-command prompts.
- [Tools](https://learnosapp.com/docs/mcp/tools): list_courses, search_lessons, list_lessons, get_lesson, get_chunk, create_flashcard, append_note, generate_presentation, create_presentation, update_presentation, replace_presentation_slides, add_presentation_slide, update_presentation_slide, get_presentation, list_presentations, get_transcript, create_audio_overview, get_audio_overview, get_audio_overview_url, get_course_progress, get_recent_activity, generate_quiz, grade_quiz, get_quiz_results.

## Blog

- [What is vibe coding? A plain-English definition (2026)](https://learnosapp.com/blog/what-is-vibe-coding): Vibe coding means building software with AI without reviewing the code. Where the term came from, how it differs from agentic engineering, and when it's safe.
- [How much does vibe coding cost? The 2026 line-item bill](https://learnosapp.com/blog/what-it-costs-to-vibe-code-an-app): A dated, line-item monthly bill for vibe coding an app at three tiers: agent plan, Vercel, Supabase, weather API and domain, plus how to cap surprise bills.
- [How to vibe code an app in 2026 with Claude Code](https://learnosapp.com/blog/how-to-vibe-code-an-app-with-claude-code): Vibe code a real app with Claude Code in 11 steps: spec, AGENTS.md, plan mode, git, Supabase with tested RLS, Vercel, and a pre-launch checklist.
- [Deploy Next.js + Supabase to Vercel: errors AI agents get wrong](https://learnosapp.com/blog/deploy-nextjs-supabase-vercel-errors): Next.js + Supabase works on localhost but breaks on Vercel? The real errors, their causes, the fixes, and the agent "fixes" you should refuse.
- [Supabase Row Level Security explained for vibe coders](https://learnosapp.com/blog/supabase-row-level-security-for-vibe-coders): What Supabase RLS does, why an exposed publishable key isn't the leak, copy-paste owner-only policies, and a second-user test that proves they work.
- [Is vibe coding safe? A pre-launch security checklist](https://learnosapp.com/blog/is-vibe-coding-safe): Yes, if you check what the agent built. The public incidents, an OWASP-mapped pre-launch checklist, and an audit prompt to paste into your agent.
- [NotebookLM vs Claude Projects vs MCP: studying with citations](https://learnosapp.com/blog/notebooklm-vs-claude-projects-vs-mcp-citations): Three ways to make an AI answer from your own notes and courses and cite the exact source, compared on citations, where your material lives and which AI you can use.
- [What is an agent harness, and why does it move the score?](https://learnosapp.com/blog/what-is-an-agent-harness): An agent harness is the code around a model that turns it into an agent. Here is what it does, and the evidence that it can matter as much as the model.
- [Claude Code vs Codex vs OpenCode vs Pi: what's different?](https://learnosapp.com/blog/claude-code-vs-codex-vs-opencode-vs-pi): The top coding-agent harnesses compared on the decisions that matter: tools, edit format, permissions, sessions, extensibility and which models they run.
- [Is RAG still worth it with long-context models? (2026)](https://learnosapp.com/blog/is-rag-still-worth-it-long-context): 1M-token windows changed the math, not the answer. When long context is enough, when retrieval wins, and what modern RAG looks like in 2026.
- [Business basics for vibe coders: what to know before you launch](https://learnosapp.com/blog/business-basics-for-vibe-coders): Five business concepts that decide whether an AI-built app makes money: market sizing, contribution margin, break-even, CAC/LTV, and pricing, with a worked example.
- [Jev vs Laya vs Von: what the benchmarks actually show](https://learnosapp.com/blog/jev-vs-laya-vs-von-benchmarks): Who ran each Jev, Laya and Von benchmark, what the accuracy, ECE and latency numbers mean, and when a frontier LLM with structured outputs is enough.
- [How data teams use Claude Code: SQL, reports, dashboards](https://learnosapp.com/blog/claude-code-for-data-teams): Four Claude Code workflows for analytics teams: warehouse SQL, recurring reports, dashboards and cleanup scripts, plus a rollout checklist for data leads.
- [How to build a tracker or calculator with Claude artifacts](https://learnosapp.com/blog/build-a-tracker-or-calculator-with-claude-artifacts): How to build a tracker or calculator with Claude artifacts, no code: a four-part brief, a test loop, sandbox limits, and safe sharing and updates.
- [How to use Claude artifacts in 2026: pick the right form](https://learnosapp.com/blog/how-to-use-claude-artifacts): Claude artifacts changed in September 2026. What an artifact is now, what's new, and a four-question test for picking a doc, deck, design, file, or page.
- [How to learn AI agents: a practical path for beginners](https://learnosapp.com/blog/how-to-learn-ai-agents): A focused way to start learning AI agents without getting lost in frameworks: learn the core loop, work through real examples, and use a 20-lesson course to build durable understanding.
- [Our first tool-use benchmark measured nothing, and that was the useful part](https://learnosapp.com/blog/what-a-tool-use-benchmark-measures): We built a benchmark to pick a model for tool routing. The first version separated zero of fifteen model pairs. Rebuilding it around disagreement produced a result we didn't expect — the accuracy ranking and the safety ranking come out almost inverted.
- [We scored Microsoft's MCP course against a quality rubric](https://learnosapp.com/blog/scoring-microsofts-mcp-course): Two Microsoft courses, 87 lessons, one six-category rubric. Both scored an A — but the interesting part is where they lost points, and what that says about scoring generated courses at all.

## Optional

- [llms-full.txt](https://learnosapp.com/llms-full.txt): All docs pages and the full MCP tool reference in one file