MCP Servers · cyanheads

obsidian-mcp-server

obsidian-mcp-server is a TypeScript-based MCP (Model Context Protocol) server that enables AI agents to read, write, search, and edit Obsidian vault notes, tags, and metadata through a structured API. It supports both STDIO and HTTP transport, with 14 tools covering note retrieval, creation, patching, and frontmatter management.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/cyanheads/obsidian-mcp-server

617

GitHub stars

Forks

TypeScript

Primary language

Apache-2.0

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	cyanheads/obsidian-mcp-server
Owner	cyanheads
Primary language	TypeScript
License	Apache-2.0 — OSI-approved
Stars	617
Forks	95
Open issues	17
Latest release	v3.2.9 (2026-06-30)
Last updated	2026-06-30
Source	https://github.com/cyanheads/obsidian-mcp-server

What obsidian-mcp-server is

Built on MCP SDK ^1.29.0 and @cyanheads/mcp-ts-core, this server exposes note operations via JSONLogic and BM25 search, surgical patch/append/replace editing at the heading/block/frontmatter level, and optional Obsidian command-palette dispatch. It integrates with the Omnisearch plugin for advanced search and supports regex-based note filtering with depth-capped recursive traversal.

Quickstart

Get the obsidian-mcp-server source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/cyanheads/obsidian-mcp-server.git2cd obsidian-mcp-server3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

AI-Driven Knowledge Base Augmentation

Agents can programmatically read, search, and enrich Obsidian vaults with structured metadata, tags, and cross-references, enabling intelligent note synthesis and knowledge graph enrichment without manual intervention.

Automated Note Creation & Maintenance

Supports atomic frontmatter management, tag reconciliation, and surgical in-place edits (via patch/append/replace), allowing automated workflows to create daily notes, update metadata, and maintain consistency across large vaults.

AI Agent Integration with Obsidian UI

Claude, Cursor, and VS Code can invoke the server to query and mutate notes, with optional command-palette dispatch for UI actions, creating a bidirectional bridge between AI agents and local knowledge management.

Implementation considerations

Requires local Obsidian instance running and accessible; the server communicates with Obsidian's API (local REST or socket-based) and is not a standalone vault engine.
MCP transport choice (STDIO vs. HTTP) affects deployment; HTTP requires network security setup (TLS, authentication, firewall rules) not provided by the server itself.
Omnisearch plugin optional but recommended for text search; without it, search modes degrade to substring or JSONLogic only—verify plugin compatibility before production use.
Frontmatter and tag mutation require clear schema agreement; the server is format-agnostic (YAML/JSON/etc.), but agents must know vault conventions to avoid data corruption.
Recursive note listing caps at 1000 entries and depth 20 by design; large vaults may require pagination or path-prefix filtering to avoid response timeouts or memory spikes.

When to avoid it — and what to weigh

Real-time Collaborative Editing Required — The server provides single-point mutation semantics without distributed locking or conflict resolution, unsuitable for multi-user simultaneous vault edits.
Obsidian Cloud Sync or Encryption Needed — This is a local vault server; it does not abstract encryption, cloud sync policies, or cross-device synchronization—those must be managed via Obsidian's native sync or third-party solutions.
Complex Plugin Ecosystem Dependency — Advanced features (Omnisearch integration for BM25 ranking) require specific community plugins to be installed; absence of those plugins silently degrades functionality rather than failing fast.
Enterprise Access Control & Audit Logging — No built-in RBAC, audit trails, or secret rotation mechanics; requires external authentication/authorization layering for regulated environments.

License & commercial use

Apache License 2.0 (Apache-2.0). A permissive OSI-approved open-source license permitting commercial use, modification, and distribution, with the requirement to include a copy of the license and notice of changes. No liability or warranty provided.

Apache-2.0 permits commercial use and redistribution. However, no warranty or support is provided by the licensor. If bundling into a commercial product or service, ensure compliance with license attribution and consider obtaining commercial support or indemnification separately. Requires review if integrating into proprietary systems or reselling as a service.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Strong
License clarity	Clear
Deployment complexity	Moderate
DEV.co fit	Good
Assessment confidence	High

Security considerations

The server exposes full note read/write/delete capabilities to any MCP client that can invoke it; no built-in RBAC, audit logging, or rate limiting. Local Obsidian API access is assumed secure (same machine or trusted network). HTTP transport requires external TLS/authentication wrapper. OBSIDIAN_ENABLE_COMMANDS=true flag opts into Obsidian command dispatch, expanding attack surface—use cautiously. No input validation or sanitization details provided; agents can construct arbitrary JSONLogic or regex patterns. Secrets in frontmatter or note content are accessible to any client.

Alternatives to consider

Obsidian Local REST API

Native Obsidian REST plugin offering direct vault access. Less feature-rich than obsidian-mcp-server (no MCP, no surgical patch tools) but avoids a separate server process. Suitable if only basic CRUD is needed and MCP is not required.

Notion API + n8n/Zapier Automation

Cloud-native alternative for knowledge base automation; offers built-in RBAC, audit logs, and hosted service. Requires vault migration and different data model, but eliminates local server operational overhead.

Custom Node.js Script + Obsidian Dataview Plugin

Lightweight in-process solution for vault queries and bulk edits. No MCP or remote access, so not suitable for AI agent integration; better suited for one-off scripted migrations or local workflows.

Software development agency

Build on obsidian-mcp-server with DEV.co software developers

obsidian-mcp-server bridges AI agents and Obsidian vaults. Install via npm, configure your transport (STDIO or HTTP), and start querying and editing notes programmatically. Review security and deployment requirements before production use.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

shadcn-ui-mcp-server mcp-server-browserbase apify-mcp-server

Related on DEV.co

Explore the category and the services that help you build with it.

MCP Servers AI Application Development API Development Custom Software Development Talk to DEV.co

obsidian-mcp-server FAQ

Does obsidian-mcp-server work without a local Obsidian app running?

No. The server requires a running Obsidian instance and communicates with its API. It is not a standalone vault engine and cannot operate on vault files directly without Obsidian's mediation.

How do I secure the HTTP transport for multi-user access?

The server itself provides no TLS, authentication, or authorization. You must layer a reverse proxy (nginx, Apache, or API gateway) with mTLS and an auth mechanism (JWT, OAuth2, etc.) in front of it.

What happens if Omnisearch plugin is not installed?

The omnisearch mode is omitted from the search_notes mode enum at startup. Queries fall back to text or JSONLogic modes. No error is raised—the feature silently degrades, so test search behavior explicitly.

Can I use this in a production CI/CD pipeline or as a serverless function?

Requires a persistent Obsidian instance, so not suitable for ephemeral serverless/FaaS. CI/CD pipelines can use it if they spin up a dedicated Obsidian sidecar, but this adds complexity and cost; consider custom scripts or native Obsidian plugin APIs for automation.

Software development & web development with DEV.co

From first prototype to production, DEV.co delivers software development services around tools like obsidian-mcp-server. Our software development agency staffs experienced software developers and web developers for custom software development, web development, integrations, and ongoing support across mcp servers and beyond.

Ready to Automate Your Knowledge Base?

Talk to DEV.co Schedule a Consultation