Stores AI memories as Markdown files for visualization in Obsidian's graph view, creating a knowledge graph with entities as nodes and relations as edges.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Obsidian Memory MCP create a person entity for John Doe who works at Tech Corp"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
MCP server that stores AI memories as Markdown files for visualization in Obsidian's graph view.
This project is a modified version of Anthropic's memory server that has been adapted for Obsidian integration. The original server stored memories in JSON format, while this version stores them as individual Markdown files with Obsidian-compatible [[link]] syntax for graph visualization.
Storage Format: Changed from JSON to individual Markdown files
Obsidian Integration: Added [[link]] syntax for relations
YAML Frontmatter: Metadata stored in frontmatter instead of JSON
File Structure: Each entity becomes a separate .md file
Removed Features: Simplified to focus on Markdown-only storage
Related MCP server: mcp-ltm
Markdown Storage: Individual .md files for each entity
Obsidian Integration: Uses [[link]] syntax for graph visualization
Knowledge Graph: Store entities, relations, and observations
Search Functionality: Query across all stored memories
YAML Frontmatter: Metadata stored in frontmatter
Each entity is stored as an individual Markdown file with:
YAML frontmatter for metadata (entityType, created, updated)
Obsidian-compatible [[links]] for relations
Organized sections for observations and relations
Example entity file (John_Doe.md):
---
entityType: person
created: 2025-07-10
updated: 2025-07-10
---
# John Doe
## Observations
- Works at Tech Corp
- Expert in TypeScript
- Lives in Tokyo
## Relations
- [[Manager of::Alice Smith]]
- [[Collaborates with::Bob Johnson]]
- [[Located in::Tokyo Office]]git clone https://github.com/YuNaga224/obsidian-memory-mcp.git
cd obsidian-memory-mcp
npm install
npm run buildThen configure in Claude Desktop:
{
"mcpServers": {
"obsidian-memory": {
"command": "node",
"args": ["/full/path/to/obsidian-memory-mcp/dist/index.js"],
"env": {
"MEMORY_DIR": "/path/to/your/obsidian/vault"
}
}
}
}Configure Claude Desktop with one of the options above
Restart Claude Desktop
Use the MCP memory tools to create entities and relations
Open Obsidian and view the graph
The knowledge graph will be visualized with:
Entity files as nodes
[[links]] as edges
Different colors for different entity types (if configured in Obsidian)
The server exposes the following tools:
create_entities: Create new entities
create_relations: Create relations between entities
add_observations: Add observations to existing entities
delete_entities: Delete entities and related data
delete_observations: Remove specific observations
delete_relations: Remove relations
read_graph: Get the entire knowledge graph
search_nodes: Search entities by query
open_nodes: Get specific entities by name
npm run watch # Watch for changes and rebuildThis project is based on Anthropic's memory server from the Model Context Protocol servers collection. We thank Anthropic for releasing the original implementation under the MIT license.
MIT License - see LICENSE file for details.
Original memory server: Copyright (c) 2024 Anthropic, PBC
Obsidian integration modifications: Copyright (c) 2025 YuNaga224
Add new observations to existing entities in the knowledge graph
Inputs: observations (required).
Create multiple new entities in the knowledge graph
Inputs: entities (required).
Create multiple new relations between entities in the knowledge graph. Relations should be in active voice
Inputs: relations (required).
Delete multiple entities and their associated relations from the knowledge graph
Inputs: entityNames (required).
Delete specific observations from entities in the knowledge graph
Inputs: deletions (required).
Delete multiple relations from the knowledge graph
Inputs: relations (required).
Open specific nodes in the knowledge graph by their names
Inputs: names (required).
Read the entire knowledge graph
Inputs: no named parameters.
Search for nodes in the knowledge graph based on a query
Inputs: query (required).
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 9 tools
Each tool has a clearly distinct purpose targeting specific operations on the knowledge graph: adding/deleting observations, creating/deleting entities/relations, opening/reading/searching nodes. No overlap exists; for example, 'create_entities' and 'create_relations' are distinct, and 'open_nodes' and 'search_nodes' serve different access functions.
All tools follow a consistent verb_noun pattern with snake_case throughout: 'add_observations', 'create_entities', 'delete_relations', etc. The naming is predictable and readable, with verbs like 'add', 'create', 'delete', 'open', 'read', and 'search' applied consistently to their respective nouns.
With 9 tools, the count is well-scoped for managing a knowledge graph, covering core operations like CRUD for entities, relations, and observations, plus graph access. Each tool earns its place without bloat, fitting a typical range of 3-15 tools for this domain.
The tool set provides strong coverage for knowledge graph management, including create, read, delete, and search operations. A minor gap exists in update capabilities (e.g., no 'update_entities' or 'update_relations'), but agents can work around this by deleting and recreating, and core workflows are well-supported.
Personal wiki and memory layer for AI assistants. Persistent, structured memory across sessions.
Persistent knowledge graph for AI agents. Remember, recall, and forget facts.
Shared long-term memory for AI agents: save and recall context as a searchable knowledge graph.
Portable AI memory shared across models and harnesses - plain markdown you own.