{ "markdown": "

\n
\n \n \n \n \n \"VibeLearn\"\n \n \n
\n

\n\n

Learn from every session. Built for Claude Code.

\n\n

\n \n \"License\"\n \n \n \"Version\"\n \n \n \"Node\"\n \n

\n\n

\n Quick Start •\n How It Works •\n VL CLI •\n Configuration •\n Privacy •\n Troubleshooting\n

\n\n

\n VibeLearn is a Claude Code plugin that silently watches what you build, extracts the concepts you encounter, and turns them into quiz questions — so you actually retain what you learn while coding.\n

\n\n---\n\n## What It Does\n\nEvery time you end a Claude Code session, VibeLearn automatically:\n\n1. **Detects your tech stack** — reads `package.json`, `pyproject.toml`, `go.mod`, etc.\n2. **Analyzes your code changes** — identifies patterns: custom hooks, API routes, TypeScript types, design patterns\n3. **Extracts learning concepts** — a single LLM call produces a session summary and a list of concepts you encountered\n4. **Generates quiz questions** — a second LLM call creates targeted questions per concept across 7 formats: `multiple_choice`, `code_reading`, `spot_the_bug`, `fill_in_blank`, `open_ended`, `true_false`, and `ordering` — selected based on concept difficulty (junior / mid / senior)\n5. **Syncs to vibelearn.dev** — your learning profile is stored securely (optional, requires `vl login`)\n\nThen run `vl quiz` to review what you learned.\n\n---\n\n## Quick Start\n\nInstall the plugin in a Claude Code session:\n\n```\n/plugin marketplace add anergcorp/vibelearn\n/plugin install vibelearn\n```\n\nRestart Claude Code. VibeLearn will start capturing learning data automatically from your next session.\n\n**Optional — connect to vibelearn.dev:**\n\n```bash\nvl login \n```\n\nGet your API key at [vibelearn.dev](https://vibelearn.dev).\n\n---\n\n## vl CLI\n\nThe `vl` command lets you review and interact with your learning data:\n\n```bash\nvl quiz # Interactive quiz — all pending questions\nvl quiz --session # Quiz questions from the last session only\n\nvl status # Sessions analyzed, top concept categories, mastery stats\nvl gaps # Concepts you haven't mastered yet (mastery < 50%)\n\nvl login # Connect to vibelearn.dev\nvl login --status # Check login status\n```\n\n**Example session:**\n\n```\n$ vl quiz\n\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n VibeLearn Quiz — 3 questions\n━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\n\nQ1/3 (intermediate) [React Server Actions]\n\n Code:\n 'use server'\n export async function createPost(data: FormData) { ... }\n\n What does the 'use server' directive tell Next.js?\n\n A) Run this function in a Web Worker\n B) Execute this function on the server, not the client\n C) Cache the function result server-side\n D) Mark the function as async-only\n\n Your answer (A/B/C/D): B\n\n ✓ Correct!\n\n Explanation: 'use server' creates a Server Action — a function that\n runs exclusively on the server. The client receives only the result.\n```\n\n### Adaptive Difficulty\n\nVibeLearn tracks your mastery per concept and adjusts difficulty automatically:\n\n- **3 correct in a row** → promoted to the next level (junior → mid → senior)\n- **1 wrong answer** → dropped back a level\n- **Follow-up questions** are inserted mid-quiz when you nail a junior question, to probe deeper understanding\n- **Open-ended answers** at senior level are evaluated by a 5-dimension rubric (accuracy, depth, trade-offs, practical reasoning, completeness)\n\n---\n\n## How It Works\n\n### 5 Lifecycle Hooks\n\n```\nSessionStart → Worker starts, session initialized\nUserPromptSubmit → Session linked to user prompt\nPostToolUse → File edits/writes/bash commands captured\nStop (Summary) → 5-step analysis pipeline runs\nSessionEnd → Session finalized\n```\n\n### Analysis Pipeline (runs at session end)\n\n```\n1. StackDetector — reads package.json/config files → vl_stack_profiles\n2. StaticAnalyzer — regex/AST patterns on code changes (hooks, routes, types…)\n3. ConceptExtractor — LLM call → session summary + concept list → vl_concepts\n4. QuizGenerator — LLM call → quiz questions per concept → vl_questions\n5. UpstreamSync — HMAC-signed POST to api.vibelearn.dev (queued offline if unavailable)\n```\n\n### Worker Service\n\nAn Express HTTP server on port **37778**, managed by Bun. Hooks talk to it over localhost. It handles all database writes and the analysis pipeline.\n\n### Database\n\nSQLite at `~/.vibelearn/vibelearn.db`. Key tables:\n\n| Table | Purpose |\n|-------|---------|\n| `vibelearn_session_summaries` | Human-readable session narratives |\n| `vl_concepts` | Extracted concepts per session |\n| `vl_questions` | Generated quiz questions |\n| `vl_quiz_attempts` | Your answers (HMAC-signed before sync) |\n| `vl_developer_profile` | Mastery score per concept |\n| `vl_stack_profiles` | Detected tech stack per session |\n| `vl_sync_queue` | Offline retry queue |\n\n---\n\n## Configuration\n\nSettings are auto-created at `~/.vibelearn/settings.json` on first run.\n\n**Key settings:**\n\n```json\n{\n \"VIBELEARN_WORKER_PORT\": \"37778\",\n \"VIBELEARN_DATA_DIR\": \"~/.vibelearn\",\n \"VIBELEARN_LOG_LEVEL\": \"INFO\",\n \"VIBELEARN_PROVIDER\": \"claude\",\n \"VIBELEARN_GEMINI_API_KEY\": \"\",\n \"VIBELEARN_OPENROUTER_API_KEY\": \"\",\n \"VIBELEARN_AUTO_SYNC\": \"true\",\n \"VIBELEARN_EXCLUDED_PROJECTS\": \"\"\n}\n```\n\n**AI Provider for Analysis**\n\nThe analysis pipeline (concept extraction + quiz generation) uses your configured LLM provider. Priority order:\n\n1. **Gemini** — set `VIBELEARN_GEMINI_API_KEY` (free tier available)\n2. **OpenRouter** — set `VIBELEARN_OPENROUTER_API_KEY`\n3. **Anthropic** — uses `ANTHROPIC_API_KEY` from environment (claude-haiku-4-5)\n\n**Excluding projects:**\n\n```json\n{\n \"VIBELEARN_EXCLUDED_PROJECTS\": \"/path/to/skip,~/personal/*\"\n}\n```\n\n---\n\n## Privacy\n\nWrap any content in `` tags to prevent it from being stored or synced:\n\n```\nPlease review my-secret-api-key: sk-... configuration\n```\n\nEverything inside `` is stripped at the hook layer before reaching the worker or database.\n\n**What is never stored:**\n- Absolute file paths (only basenames are sent upstream)\n- Raw file contents (only short snippets from the analysis)\n- Full user prompts\n\n**Anti-tamper:** Quiz attempt records are HMAC-signed with your API key before syncing. The server recomputes your streak from accepted attempt records — local SQLite data cannot be used to fake progress.\n\n---\n\n## System Requirements\n\n- **Node.js**: 18.0.0 or higher\n- **Claude Code**: Latest version with plugin support\n- **Bun**: JavaScript runtime (auto-installed if missing)\n\n---\n\n## Windows Notes\n\nIf you see `npm : The term 'npm' is not recognized`:\n\nInstall [Node.js](https://nodejs.org) and restart your terminal. Bun is auto-installed by the plugin setup script.\n\n---\n\n## Build\n\n```bash\nnpm install\nnpm run build-and-sync # Build + sync to marketplace + restart worker\n```\n\nBuilt outputs land in `plugin/scripts/`:\n- `worker-service.cjs` — the worker daemon\n- `mcp-server.cjs` — MCP tools\n- `vl-cli.cjs` — the `vl` binary\n\n---\n\n## Contributing\n\n1. Fork the repository\n2. Create a feature branch\n3. Make your changes with tests\n4. Update documentation\n5. Submit a Pull Request\n\n---\n\n## License\n\nGNU Affero General Public License v3.0 (AGPL-3.0).\n\nSee the [LICENSE](LICENSE) file for full details.\n\n---\n\n## Support\n\n- **Issues**: [GitHub Issues](https://github.com/anergcorp/vibelearn/issues)\n- **Website**: [vibelearn.dev](https://vibelearn.dev)\n- **Repository**: [github.com/anergcorp/vibelearn](https://github.com/anergcorp/vibelearn)\n\n---\n\n**Built with Claude Agent SDK** | **Powered by Claude Code** | **Made with TypeScript**\n", "bytes": 8691, "sha": "4ecb0299c386504e759c16a06ca624c1e3a5d94273e058c645e95ff0c298e38d", "repo_slug": "anergcorp/vibelearn", "fonte": "repo", "truncated": false, "api": "https://agentalog.com/api/listings/plg_anergcorp_vibelearn_vibelearn_aec9b7e1/readme" }