Obsidian Echobrain — AI skill for Claude Code
Make Obsidian the Local Brain for AI Agents.
How to install Obsidian Echobrain
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open dxer/obsidian-echobrain and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Obsidian Echobrain does
Make Obsidian the Local Brain for AI Agents. Embedded MCP Server (HTTP/SSE) with Graph-RAG for WorkBuddy, Cursor & Claude Desktop.
Alternatives in AI
- Nocturne Memory — A lightweight, rollbackable, and visual Long-Term Memory Server for MCP Agents 1.3k ★
- Kindly Web Search MCP Server — Kindly Web Search MCP Server: Web search + robust content retrieval for AI coding tools (Claude Code, Codex, C 377 ★
- Engraph — Local knowledge graph for AI agents 124 ★
README
EchoBrain Local
**让 Obsidian 成为 Agent 的本地大脑。** *Make Obsidian the Local Brain for AI Agents.*
[](LICENSE) [](https://modelcontextprotocol.io/) [](https://obsidian.md/) [](https://www.typescriptlang.org/)
EchoBrain Local 是一个运行在 Obsidian 内部的本地优先(Local-First)模型上下文协议(Model Context Protocol, MCP)插件。
它的核心不是“做一个知识库”,也不是“做一个 MCP 工具集合”,而是: **让外部 AI Agent 能够真正读取、理解、关联和沉淀你在 Obsidian 中的个人知识。**
无需配置 Node.js 运行时或外部矢量数据库,只需在 Obsidian 内部开启插件,即可让 **WorkBuddy (腾讯)**、**Cursor**、**Claude Desktop** 等现代 Agent 拥有你的长期经验记忆。
核心心智模型 (The Mental Model)
┌─────────────────────────────────────────────────────────────┐
│ Obsidian = Agent 的长期记忆 (Long-Term Memory) │
│ EchoBrain = 连接记忆与 Agent 的神经中枢 (Nervous System) │
│ MCP = Agent 调度记忆的突触接口 (Standard Protocol) │
│ Inbox = Agent 写入新记忆的安全隔离区 (Safe Entry) │
└─────────────────────────────────────────────────────────────┘
- 记忆主体在本地:知识库始终由你完全掌控,零数据上云,断网完全可用。
- 神经中枢单进程:利用 Obsidian 原生渲染进程与文件缓存,无外部 Node 进程膨胀。
- 单向真理原则:外部 Agent 只能检索已有笔记,新沉淀严格落入
Inbox/收件箱,永不篡改历史文档。
四维闭环:读、懂、联、沉淀
┌────────┐ ┌────────┐ ┌────────┐ ┌────────┐
│ 1. 读 │ → │ 2. 懂 │ → │ 3. 联 │ → │ 4. 沉淀│
└────────┘ └────────┘ └────────┘ └────────┘
精准词频 语义向量 图谱拓扑 安全收件
(BM25/分词) (ONNX/WASM) (PageRank) (Inbox)
- 读 (Precision Read):自研代码标识符切词与中文字词滑动窗口(2-gram / 3-gram),毫秒级精准命中函数名、配置项、报错码及专有名词。
- 懂 (Semantic Understand):三级按需嵌入架构——零开销纯词频模式、按需加载 40MB 本地 WebAssembly ONNX 模型(
bge-small-zh-v1.5),或接入任意兼容 OpenAI / Ollama 的向量 API,穿透字面差异识别模糊意图。 - 联 (Topological Relate):直接挂载 Obsidian 原生 `met
Related Skills
Kuzu Memory
Lightweight, embedded graph-based memory system for AI applications. Fast (<3ms recall), offline-first, with M
Token Watcher
⏱ Token Watcher — local real-time token usage & quota dashboard for AI coding agents (Claude Code, ccmr, Codex
Brainstem MCP
Your Obsidian vault as Claude's second brain: self-hosted MCP connector for claude.ai, Claude mobile/Desktop a
Graph Query MCP
FastMCP server for surgical queries against a vault knowledge graph. NetworkX node-link JSON. Companion to gra
MeMex Zero RAG
The Karpathy LLM Wiki, production-ready. Zero-RAG personal knowledge base with MCP server, multi-agent support
Memento
A shared brain for your AI coding tools — collects conversations, plans and memory from Claude Code / Cursor /
Related Agents
Ehs Local App
Local application security specialist for the Ethical Hacker Squad. Reviews command-line tools, desktop applic
Optimizer MCP
Specialized agent for optimizing Claude Code MCP server configurations. Expert in .mcp.json structure, server
Brain Retrieval Engineer
Retrieval engineer — six retrieval strategies (hybrid, graph-enhanced vector, entity-anchored local, guarded T