{
  "markdown": "<p align=\"center\">\n  <img src=\"docs/assets/banner.svg\" alt=\"Alvarmethod — probe, plan, teach\" width=\"960\" />\n</p>\n\n<p align=\"center\">\n  <strong>One teacher. One mind.</strong><br/>\n  Portable agent skills that run Eero Alvar’s AI learning loop.\n</p>\n\n<p align=\"center\">\n  <a href=\"https://github.com/vasanthsreeram/Alvarmethod/stargazers\"><img src=\"https://img.shields.io/github/stars/vasanthsreeram/Alvarmethod?style=flat-square&color=c9a227\" alt=\"stars\" /></a>\n  <a href=\"LICENSE\"><img src=\"https://img.shields.io/badge/license-MIT-2d6a4f?style=flat-square\" alt=\"MIT\" /></a>\n  <a href=\"https://agentskills.io\"><img src=\"https://img.shields.io/badge/Agent%20Skills-SKILL.md-111827?style=flat-square\" alt=\"Agent Skills\" /></a>\n  <a href=\"https://youtu.be/kzcI5F4tGiU\"><img src=\"https://img.shields.io/badge/method-Eero%20Alvar-b45309?style=flat-square\" alt=\"Eero Alvar\" /></a>\n</p>\n\n```bash\nnpx skills add vasanthsreeram/Alvarmethod -g --all\n```\n\nThat’s the install. Works in **Claude Code**, **Codex**, **Grok**, **Pi**, **OpenCode**, Cursor, and the rest of the [skills CLI](https://skills.sh) agents.\n\nThen open a **learning folder** (not this repo) and say:\n\n```text\n/teach I want a solid introduction to <topic>\n```\n\nYou should get a **quiz picker** (not A/B/C/D in chat) → a mermaid DAG → **one** reasoning step → a lock-in quiz.\n\n---\n\n## Quick install\n\n### `npx skills` (preferred)\n\nThe [open skills CLI](https://github.com/vercel-labs/skills). Discovers every `skills/*/SKILL.md` in this repo.\n\n```bash\n# global, every detected agent, no prompts\nnpx skills add vasanthsreeram/Alvarmethod -g --all\n\n# list what’s in the pack\nnpx skills add vasanthsreeram/Alvarmethod --list\n\n# one skill\nnpx skills add vasanthsreeram/Alvarmethod --skill teach -g -y\n\n# one harness\nnpx skills add vasanthsreeram/Alvarmethod -g -y -a claude-code -a grok -a pi -a opencode -a codex\n```\n\nUpdate later:\n\n```bash\nnpx skills update -g\n```\n\n### `npx` this repo\n\nNo extra CLI — runs [`install.sh`](install.sh) from GitHub:\n\n```bash\nnpx github:vasanthsreeram/Alvarmethod\nnpx github:vasanthsreeram/Alvarmethod --list\nnpx github:vasanthsreeram/Alvarmethod --claude --grok\n```\n\n### Clone\n\n```bash\ngit clone https://github.com/vasanthsreeram/Alvarmethod.git\ncd Alvarmethod\n./install.sh          # all known agent dirs\n./install.sh --project\n./install.sh --uninstall\n```\n\n| Agent | Global path |\n|-------|-------------|\n| Codex | `~/.codex/skills` + `~/.agents/skills` |\n| Claude Code | `~/.claude/skills` |\n| Grok | `~/.grok/skills` |\n| Pi | `~/.pi/agent/skills` |\n| OpenCode | `~/.config/opencode/skills` |\n\n---\n\n## The method\n\nCredit: **[Eero Alvar — *How I Use AI to Learn Things*](https://youtu.be/kzcI5F4tGiU)** (2026-08-14). This pack implements that loop as Agent Skills. It is not his original Pi quiz / Obsidian setup. Captions: [`source/eero-alvar-how-i-use-ai-to-learn-things.md`](source/eero-alvar-how-i-use-ai-to-learn-things.md).\n\nMany-to-many learning leaks:\n\n| Direction | Waste |\n|-----------|--------|\n| One outlet → many students | Never fitted to *your* edge |\n| One student → many outlets | Switching cost + the brain hedges on sources it does not trust |\n\nFix: **one interface, many sources.** Trust is engineered (verify). Struggle stays in the material. The system eats logistics.\n\n```mermaid\nflowchart LR\n  A[Probe<br/>edge of this mind] --> B[Plan<br/>mermaid DAG]\n  B --> C[Teach<br/>one node]\n  C --> D[Lock-in quiz]\n  D -->|pass| C\n  D -->|fail| E[Insert prereq]\n  E --> C\n```\n\n1. **Probe** — graded MCQ in the harness picker, start wide, binary-search every strand\n2. **Plan** — mermaid DAG for *this* mind, shown *before* teaching\n3. **Teach** — one reasoning step, picture if needed, quiz, next node\n\n---\n\n## Skills\n\n| Skill | Job |\n|-------|-----|\n| `teach` | Full loop: probe → plan → one-step teaching |\n| `probe` | Understanding map only |\n| `learn-profile` | Write `.alvar/LEARNER.md` |\n| `learn-visual` | One SVG, look at it, fix it |\n| `learn-verify` | Fact-check a claim before it is taught as fact |\n\nQuizzes **must** use the native picker:\n\n| Agent | Tool |\n|-------|------|\n| Grok / Codex | `ask_user_question` |\n| Claude Code | `AskUserQuestion` |\n| OpenCode | `question` |\n| Pi | `quiz` / `ask_user` / `askUserQuestion` |\n\nIf the agent pastes A/B/C/D, say **“use the quiz tool.”** If it dumps a textbook, say **“one node only.”**\n\nSession files land in the learning folder:\n\n```text\n.alvar/LEARNER.md\n.alvar/maps/<topic>.md\n.alvar/sessions/<date>-<topic>.md\n.alvar/visuals/…\n```\n\nPoint Obsidian at `.alvar/` if you want LaTeX preview.\n\n---\n\n## Contributing\n\nSee [CONTRIBUTING.md](CONTRIBUTING.md). Edit skills **here**, then reinstall. Don’t patch the copies under `~/.claude` / `~/.grok`.\n\n## License\n\nMIT for the skills, installer, and docs.\n\nThe video and spoken method are **Eero Alvar’s**. Watch the original: [youtu.be/kzcI5F4tGiU](https://youtu.be/kzcI5F4tGiU).\n",
  "bytes": 4890,
  "sha": "08d48ce5c768f8ed6cb88a722b71e1040dd7b9cf2051cfb852fbd845f2f632c0",
  "repo_slug": "vasanthsreeram/alvarmethod",
  "fonte": "repo",
  "truncated": false,
  "api": "https://api.agentalog.com/api/listings/skl_vasanthsreeram_alvarmethod_learn_verify_f3a288f7/readme"
}