{
  "markdown": "# learn\n\nResearch any topic online and create comprehensive learning guides with RAG-optimized indexes for your AI agents.\n\n## Why\n\nAI agents work better when they have curated, pre-researched knowledge to draw from instead of searching the web on every question. `/learn` builds that knowledge base systematically - gathering sources, scoring them for quality, and synthesizing structured guides that agents can reference instantly.\n\nUse cases:\n\n- Learning a new technology before starting implementation\n- Building a shared knowledge base across a team's AI tools\n- Creating authoritative reference material from scattered online sources\n- Producing guides that work as RAG context for Claude Code, OpenCode, and Codex\n\n## Installation\n\n```bash\nagentsys install learn\n```\n\nRequires [agentsys](https://github.com/agent-sh/agentsys) to be set up in your project.\n\n## Quick Start\n\n```\n/learn react hooks\n```\n\nThis searches the web for ~20 sources on React hooks, scores each source for authority and depth, fetches the top results, and writes a synthesized guide to `agent-knowledge/react-hooks.md` with a companion `resources/react-hooks-sources.json` containing full source metadata.\n\n## How It Works\n\nThe learn skill follows a six-stage methodology:\n\n1. **Progressive discovery** - Funnel approach: broad queries for landscape mapping, focused queries for core content, deep queries for advanced material. Avoids noise from dumping all queries at once.\n\n2. **Quality scoring** - Each source is scored on a 100-point scale across five dimensions: authority (3x weight), recency (2x), depth (2x), examples (2x), and uniqueness (1x). Official docs score highest; undated blog posts score lowest.\n\n3. **Just-in-time extraction** - Only high-scoring sources get fetched. Summaries and key insights are extracted - never full content. This keeps token usage predictable and respects copyright.\n\n4. **Synthesis** - A structured learning guide is generated with prerequisites, core concepts, code examples, common pitfalls, best practices, and further reading. Content is cross-referenced across sources, not copied from any single one.\n\n5. **RAG index** - The master index (`agent-knowledge/CLAUDE.md` and `AGENTS.md`) is updated with the new topic, trigger phrases, and keyword mappings so agents can find relevant guides automatically.\n\n6. **Enhancement** - Runs `enhance:enhance-docs` and `enhance:enhance-prompts` on the output to improve RAG retrieval quality. Skip with `--no-enhance`.\n\n## Usage\n\n```bash\n# Default depth (20 sources)\n/learn recursion\n\n# Deep research (40 sources)\n/learn kubernetes networking --depth=deep\n\n# Quick overview (10 sources)\n/learn python decorators --depth=brief\n\n# Skip enhancement pass\n/learn typescript generics --no-enhance\n```\n\n### Depth Levels\n\n| Level | Sources | When to Use |\n|-------|---------|-------------|\n| `brief` | 10 | Quick overview, time-sensitive topics |\n| `medium` | 20 | Balanced coverage (default) |\n| `deep` | 40 | Comprehensive research, complex topics |\n\n### Output Files\n\nEach run creates or updates:\n\n```\nagent-knowledge/\n  CLAUDE.md                       # Master index (updated)\n  AGENTS.md                       # Master index for OpenCode/Codex (updated)\n  <topic-slug>.md                 # Synthesized learning guide\n  resources/\n    <topic-slug>-sources.json     # Source metadata with quality scores\n```\n\n### Existing Topics\n\nIf a guide already exists for the topic, you are prompted to either update the existing guide with new sources or start fresh.\n\n## Architecture\n\n| Component | Type | Model | Role |\n|-----------|------|-------|------|\n| `learn` | command | - | Entry point, argument parsing |\n| `learn-agent` | agent | sonnet | Research coordination, web search, synthesis |\n| `learn` | skill | - | Research methodology, scoring rubric, templates |\n\n## Requirements\n\n- [agentsys](https://github.com/agent-sh/agentsys) runtime\n- Web access (WebSearch and WebFetch tools)\n- An `agent-knowledge/` directory in the workspace (created automatically)\n\n## Related Plugins\n\n- [agent-knowledge](https://github.com/agent-sh/agent-knowledge) - Where guides are stored; contains existing research\n- [enhance](https://github.com/agent-sh/enhance) - Post-processing for RAG optimization\n- [consult](https://github.com/agent-sh/consult) - For getting a second opinion on specific questions instead of building a full guide\n\n## License\n\nMIT\n",
  "bytes": 4400,
  "sha": "2c3a122eb0b0633765dbef060b83faa460f8aa4ca3b97d75a7f065881a062cc7",
  "repo_slug": "agent-sh/learn",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_agent_sh_learn_learn_48a5689f/readme"
}