Engraph
Local knowledge graph for AI agents. Hybrid search + MCP server for Obsidian vaults.
What is Engraph?
Engraph is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to local knowledge graph for ai agents. hybrid search + mcp server for obsidian vaults.
Local knowledge graph for AI agents. Hybrid search + MCP server for Obsidian vaults.
This server falls under the Knowledge & Memory category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- Local knowledge graph for AI agents. Hybrid search + MCP ser
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx engraphConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Engraph
Engraph turns your Obsidian markdown vault into a fully searchable knowledge graph that any AI agent can query through MCP, HTTP REST, or ChatGPT Actions — all running locally with no API keys and no cloud dependency. It combines five retrieval lanes (semantic embeddings, BM25 full-text, wikilink graph expansion, cross-encoder reranking, and temporal scoring) fused via Reciprocal Rank Fusion, and exposes 25 MCP tools covering search, note reading and editing, frontmatter mutations, context bundles, vault health diagnostics, and note creation. Developers and researchers who want their AI coding assistant to understand the full context of their notes and follow wikilink connections will find engraph especially useful.
Prerequisites
- An Obsidian vault with markdown notes on your local machine
- macOS (arm64) or Linux (x86_64) — pre-built binaries available, or Rust + CMake to build from source
- Approximately 300 MB of free disk space for the mandatory GGUF embedding model (downloaded on first index)
- An MCP-compatible client such as Claude Desktop or Claude Code
Install engraph
Install via Homebrew on macOS, download a pre-built binary from the GitHub releases page, or build from source with Cargo.
# macOS via Homebrew
brew install devwhodevs/tap/engraph
# From source (requires CMake)
cargo install --git https://github.com/devwhodevs/engraphIndex your Obsidian vault
Run the index command pointing at your vault directory. On first run engraph downloads an embedding model (~300 MB). Subsequent runs are incremental — only changed files are re-embedded.
engraph index ~/path/to/your/obsidian-vaultTest search from the command line
Verify the index is working by running a search query. Results include the source file, heading, and a relevance score.
engraph search "how does the auth system work"Configure your MCP client
Add engraph to your MCP client settings. For Claude Code, edit ~/.claude/settings.json. For Claude Desktop, edit claude_desktop_config.json.
{
"mcpServers": {
"engraph": {
"command": "engraph",
"args": ["serve"]
}
}
}Restart your MCP client and start querying your vault
Restart Claude Desktop or Claude Code. Engraph will start serving the 25 MCP tools automatically. The file watcher keeps the index fresh as you edit notes in Obsidian.
Engraph Examples
Client configuration
Claude Code settings.json configuration for engraph.
{
"mcpServers": {
"engraph": {
"command": "engraph",
"args": ["serve"]
}
}
}Prompts to try
Example queries that leverage engraph's hybrid search and knowledge graph capabilities.
- "Search my vault for notes about the authentication architecture"
- "What did I write about project planning last month?"
- "Find all notes linked to my API Design note"
- "Show me notes tagged #todo that mention performance"
- "Create a new note summarizing what I know about OAuth 2.0"Troubleshooting Engraph
First index run is slow or appears to stall
On the first run engraph downloads a ~300 MB GGUF embedding model. This is a one-time download. Subsequent incremental runs are much faster. Metal GPU acceleration is used automatically on macOS Apple Silicon.
Search returns no results after indexing
Verify the vault path passed to `engraph index` is correct and contains markdown (.md) files. You can re-run the index command to force a refresh. Check that the vault path in `engraph serve` matches where the index was built.
MCP client cannot connect to engraph
Ensure `engraph` is in your PATH by running `which engraph` in a terminal. If installed via Homebrew, confirm the Homebrew bin directory is in your shell's PATH, then restart your MCP client.
Frequently Asked Questions about Engraph
What is Engraph?
Engraph is a Model Context Protocol (MCP) server that local knowledge graph for ai agents. hybrid search + mcp server for obsidian vaults. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Engraph?
Follow the installation instructions on the Engraph GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Engraph?
Engraph works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Engraph free to use?
Yes, Engraph is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
Engraph Alternatives — Similar Knowledge & Memory Servers
Looking for alternatives to Engraph? Here are other popular knowledge & memory servers you can use with Claude, Cursor, and VS Code.
MemPalace
★ 52.6kA local AI memory system that stores all conversations verbatim and organizes them into navigable structures. It provides 19 MCP tools for AI assistants to search and retrieve past decisions, debugging sessions, and architecture debates automatically
Kratos
★ 25.7k🏛️ Memory System for AI Coding Tools - Never explain your codebase again. MCP server with perfect project isolation, 95.8% context accuracy, and the Four Pillars Framework.
Context Mode
★ 15.4kAn MCP server that preserves LLM context by intercepting large data outputs and returning only concise summaries or relevant sections. It enables efficient sandboxed code execution, file processing, and documentation indexing across multiple programm
Memu
★ 13.7kMemory for 24/7 proactive agents like OpenClaw.
MemOS
★ 9.3kMemOS (Memory Operating System) is a memory management operating system designed for AI applications. Its goal is: to enable your AI system to have long-term memory like a human, not only remembering what users have said but also actively invoking, u
Everos
★ 5.4kBuild, evaluate, and integrate long-term memory for self-evolving agents.
Browse More Knowledge & Memory MCP Servers
Explore all knowledge & memory servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Engraph in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Engraph?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.