跳转到主要内容
此内容尚未提供您的语言版本,正在以英文显示。

LLM Wiki

插件 已验证 活跃

Turn Claude Code + Obsidian into a second brain. The LLM incrementally ingests sources into a persistent, interlinked markdown wiki — building entity/concept/source pages, flagging contradictions, maintaining an index and log. Knowledge compounds instead of being re-derived by RAG on every query. Inspired by Karpathy's LLM Wiki gist. Ships SKILL, 3 sub-agents, 5 slash commands, 8 Python tools (stdlib only), full vault templates, and cross-tool compatibility (Claude Code, Codex CLI, Cursor, Antigravity, OpenCode, Gemini CLI).

1 个 Skill 0 个 MCP
目的

To provide a disciplined and compounding knowledge management system, turning an LLM and Obsidian into a 'second brain' where knowledge accumulates and is actively maintained.

功能

  • Incremental source ingestion into a persistent wiki
  • Automatic entity/concept page creation and updates
  • Flagging and management of contradictions
  • Maintenance of an interlinked knowledge graph and log
  • Compounding knowledge over time, reducing RAG reliance

使用场景

  • Building a personal knowledge base for research, journaling, or self-improvement.
  • Creating a companion wiki for books or research papers.
  • Developing an internal team knowledge base from meeting notes and articles.
  • Tracking evolving theses or complex topics over time.

非目标

  • Replacing all RAG workflows (it complements RAG at scale).
  • One-shot Q&A over fixed documents (use RAG for this).
  • Managing knowledge without user curation or Obsidian integration.

工作流

  1. Initialize vault (`/wiki-init`)
  2. Add source file to `raw/` directory
  3. Ingest source (`/wiki-ingest <path>`)
  4. Discuss and confirm LLM's proposed changes
  5. Query the wiki (`/wiki-query <question>`)
  6. Optionally file query answers back into wiki
  7. Periodically run health checks (`/wiki-lint`)
  8. Review wiki log (`/wiki-log`)

实践

  • Knowledge compounding
  • Incremental knowledge base maintenance
  • LLM-assisted wiki curation
  • Source integrity and contradiction flagging

先决条件

  • Obsidian installed and configured.
  • Python 3.x with standard library.
  • Claude Code or compatible LLM CLI (Codex, Cursor, etc.).

安装

请先添加 Marketplace

/plugin marketplace add alirezarezvani/claude-skills
/plugin install llm-wiki@claude-code-skills

质量评分

已验证
99 /100
1 day ago 分析

信任信号

最近提交1 day ago
星标14.6k
许可证MIT
状态
查看源代码