zxfccmm4/Obsidian-OpenCode-Knowledge

What it solves

This project provides a local, automated system for managing knowledge within the Obsidian note-taking app. It solves the problem of manually organizing, summarizing, and maintaining notes by using AI agents to process raw information, extract insights from social media, and keep a knowledge base healthy and structured.

How it works

The system acts as a bridge between Obsidian and an AI agent (such as OpenCode, Claude Code, or Codex). Users interact with the agent directly within their notes to ingest content (PDFs, articles, screenshots), capture social media data, or perform "health checks" on their knowledge base. The project uses a set of pre-configured AI skills and rules that guide the agent on how to categorize, summarize, and link information, all while keeping the actual note files stored locally on the user's machine.

Who it’s for

It is designed for non-technical users who want to build a "second brain" using AI without needing to write code. It is particularly useful for researchers, content creators, and knowledge workers who consume large amounts of information from various sources and want to automate the process of turning that information into structured, searchable notes.

Highlights

• Automated Ingestion: Automatically converts raw inputs like articles, PDFs, and screenshots into structured Markdown notes.
• Social Media Integration: Uses OpenCLI to capture and digest content from platforms like X (Twitter), Xiaohongshu, and Douyin.
• Knowledge Maintenance: Includes an AI-powered "health check" feature to identify dead links, duplicate content, and isolated pages.
• Local-First Privacy: Notes are stored locally on your device; only the content being processed is sent to your chosen AI model provider.
• One-Click Deployment: Provides simple scripts to set up the environment and configure the AI agent of your choice.