obsidian-mcp-server
MCP ServerFreeObsidian Knowledge-Management MCP (Model Context Protocol) server that enables AI agents and development tools to interact with an Obsidian vault. It provides a comprehensive suite of tools for reading, writing, searching, and managing notes, tags, and frontmatter, acting as a bridge to the Obsidian
Capabilities13 decomposed
mcp-compliant vault bridging via stdio and http transports
Medium confidenceImplements dual-transport MCP server architecture (stdio for local CLI/IDE integration, HTTP for remote agents) that translates MCP protocol messages into Obsidian Local REST API calls. Uses @modelcontextprotocol/sdk with a layered transport abstraction pattern, maintaining separate Server instances per transport mode while sharing a unified service layer for vault operations. Stdio transport creates persistent process-based communication for tools like Claude Desktop; HTTP transport exposes the same MCP tools over REST with configurable CORS and authentication.
Dual-transport architecture with shared service layer enables both local (stdio) and remote (HTTP) MCP clients to access the same vault operations without code duplication. Uses @modelcontextprotocol/sdk's transport abstraction pattern to decouple protocol handling from business logic, allowing transport-agnostic tool definitions.
Supports both local IDE integration (stdio) and remote agent access (HTTP) in a single server, whereas most MCP implementations are transport-specific or require separate deployments.
vault-aware note reading with metadata extraction
Medium confidenceImplements obsidian_read_note tool that retrieves file content and YAML frontmatter metadata via the Obsidian REST API's /vault/read endpoint, with automatic parsing of frontmatter using YAML deserialization. Supports reading by file path with optional directory filtering and returns structured output containing raw content, parsed frontmatter object, and file metadata (creation/modification timestamps). Uses schema validation to ensure path safety and prevent directory traversal attacks.
Combines content retrieval with automatic YAML frontmatter deserialization and returns structured metadata alongside raw content, enabling agents to reason about both note text and its semantic properties (tags, custom fields) in a single call. Uses Obsidian's REST API /vault/read endpoint rather than direct file system access, ensuring consistency with Obsidian's internal state.
Provides structured frontmatter parsing out-of-the-box (unlike raw file readers), and integrates with Obsidian's REST API for consistency, whereas direct file system access could read stale or partially-written content.
input sanitization and schema validation with redos prevention
Medium confidenceImplements multi-layer input validation using JSON Schema validation for all MCP tool parameters, regex pattern analysis to detect ReDoS vulnerabilities, and path traversal prevention via path normalization and allowlist checking. Validates file paths against vault root to prevent directory traversal attacks, sanitizes regex patterns before passing to Obsidian's search engine, and enforces content size limits. Uses zod or similar schema validation library with custom validators for domain-specific constraints.
Combines JSON Schema validation, regex ReDoS detection, and path traversal prevention in a unified validation layer that runs before any Obsidian REST API calls. Uses heuristic-based ReDoS detection to identify potentially dangerous patterns without executing them.
Multi-layer validation (schema + regex analysis + path checking) provides defense-in-depth, whereas single-layer validation may miss edge cases. ReDoS detection prevents performance attacks without requiring regex execution.
vault state caching with invalidation strategy
Medium confidenceImplements VaultCacheService that maintains an in-memory cache of frequently accessed vault metadata (file listings, search results, frontmatter) with configurable TTL-based invalidation. Supports manual cache invalidation on write operations (note updates, deletions) to maintain consistency. Uses LRU eviction policy to prevent unbounded memory growth. Cache keys are based on operation parameters (path, search query, etc.) enabling fine-grained invalidation.
Implements LRU-based in-memory caching with TTL invalidation and manual invalidation on write operations, enabling fast repeated access to vault data without polling Obsidian REST API. Cache keys are based on operation parameters enabling fine-grained invalidation.
In-memory caching provides sub-millisecond latency for cached queries (vs 50-200ms for REST API calls), with automatic TTL-based invalidation ensuring eventual consistency. Manual invalidation on writes prevents serving stale data after updates.
modular tool registration and extensibility framework
Medium confidenceImplements tool registration system where each MCP tool (obsidian_read_note, obsidian_update_note, etc.) is defined as a separate module with standardized interface: name, description, input schema, and handler function. Tools are registered with the MCP server via a registry pattern, enabling dynamic tool discovery and addition of custom tools without modifying core server code. Each tool module exports its schema and handler independently, allowing tools to be tested, versioned, and deployed separately.
Uses modular tool registration pattern where each tool is a separate module with standardized interface, enabling independent testing, versioning, and deployment. Tools are registered dynamically at server startup via a registry, allowing custom tools to be added without modifying core code.
Modular architecture enables independent tool development and testing (unlike monolithic tool implementations), supports dynamic registration enabling plugin-like extensibility, and allows tools to be versioned and deployed separately.
vault-wide full-text search with regex and content filtering
Medium confidenceImplements obsidian_global_search tool that executes vault-wide content searches via Obsidian REST API's /search/simple endpoint, supporting both plain-text and regex pattern matching with optional result filtering by file type, path prefix, or tag. Returns ranked search results with file paths, matching line snippets, and match positions. Uses schema validation to sanitize regex patterns and prevent ReDoS attacks, with configurable result limits to prevent memory exhaustion.
Leverages Obsidian's native search index and regex engine via REST API, enabling vault-wide searches without re-indexing or maintaining a separate search backend. Supports both plain-text and regex patterns with configurable result filtering and limits, integrated into the MCP tool schema with input validation to prevent ReDoS attacks.
Uses Obsidian's built-in search index (faster than external indexing) and integrates directly with Obsidian's regex dialect, whereas external search tools would require maintaining a separate index and may have different regex semantics.
content-aware note updating with append/prepend/overwrite modes
Medium confidenceImplements obsidian_update_note tool that modifies note content via Obsidian REST API's /vault/modify endpoint with three distinct modes: append (add content to end), prepend (add content to start), or overwrite (replace entire content). Preserves YAML frontmatter during updates and supports atomic multi-line insertions. Uses schema validation to prevent path traversal and enforces content size limits to prevent vault corruption.
Provides three distinct update modes (append/prepend/overwrite) in a single tool with automatic frontmatter preservation, enabling flexible content modification patterns without requiring separate tools. Uses Obsidian's /vault/modify endpoint for atomic updates, ensuring consistency with Obsidian's internal state and file watchers.
Supports append/prepend modes natively (unlike simple file overwrite tools), preserves frontmatter automatically, and integrates with Obsidian's file system watchers, whereas direct file writes could corrupt frontmatter or trigger race conditions.
regex-based search and replace with validation
Medium confidenceImplements obsidian_search_replace tool that performs targeted text and regex replacements within a single note via Obsidian REST API's /vault/modify endpoint with search pattern validation. Supports both literal string and regex pattern matching with optional case-insensitive and global flags. Validates regex patterns before execution to prevent ReDoS attacks, and returns match count and preview of changes before applying. Uses atomic updates to ensure consistency.
Integrates regex pattern validation with atomic replacements via Obsidian's REST API, preventing ReDoS attacks while supporting both literal and regex patterns. Returns match count and change preview before applying, enabling safer bulk operations than raw file replacement.
Validates regex patterns server-side to prevent ReDoS attacks (unlike naive regex tools), integrates with Obsidian's file system for consistency, and supports both literal and regex patterns in a single tool.
yaml frontmatter crud operations with schema validation
Medium confidenceImplements obsidian_manage_frontmatter tool that performs create, read, update, and delete operations on YAML frontmatter fields via Obsidian REST API. Parses existing frontmatter, applies field-level modifications (add, update, delete, or replace entire frontmatter), and re-serializes to YAML with proper formatting. Uses schema validation to ensure valid YAML structure and prevents injection attacks. Supports nested field access via dot notation (e.g., 'metadata.author').
Provides unified CRUD interface for frontmatter fields with dot-notation support for nested access, automatic YAML serialization/deserialization, and schema validation to prevent injection attacks. Integrates with Obsidian's REST API to ensure consistency with Obsidian's frontmatter handling.
Supports all CRUD operations in a single tool (unlike separate read/write tools), handles YAML serialization automatically, and validates input to prevent injection, whereas raw file editing could corrupt YAML syntax.
tag management across frontmatter and inline syntax
Medium confidenceImplements obsidian_manage_tags tool that adds, removes, and lists tags in both YAML frontmatter ('tags' field) and inline Obsidian syntax (#tag). Automatically detects tag location (frontmatter vs inline) and applies modifications to the appropriate location. Supports bulk tag operations and tag normalization (lowercase, slug format). Uses schema validation to ensure valid tag format and prevents duplicate tags.
Manages tags across both YAML frontmatter and inline Obsidian syntax in a single tool, with automatic location detection and optional normalization. Uses regex-based inline tag detection integrated with Obsidian's REST API for consistency.
Handles both frontmatter and inline tags without requiring separate tools, supports bulk operations, and integrates with Obsidian's tag system, whereas manual tag editing could miss inline tags or create inconsistencies.
directory listing with recursive traversal and filtering
Medium confidenceImplements obsidian_list_notes tool that enumerates vault files via Obsidian REST API's /vault/list endpoint with support for recursive directory traversal, file type filtering (markdown, images, etc.), and path prefix filtering. Returns structured file metadata including paths, file types, creation/modification timestamps, and optional file sizes. Uses schema validation to prevent directory traversal attacks and supports pagination for large directories.
Provides recursive directory listing with file type and path filtering in a single tool, integrated with Obsidian's REST API for consistency. Returns structured metadata (timestamps, file types) enabling agents to understand vault structure and make informed decisions about which files to process.
Uses Obsidian's REST API for consistency (unlike direct file system access), supports filtering and pagination to handle large vaults efficiently, and returns structured metadata for downstream processing.
atomic file deletion with safety validation
Medium confidenceImplements obsidian_delete_note tool that removes files from the vault via Obsidian REST API's /vault/delete endpoint with optional safety checks (confirmation flag, path validation). Supports deletion of individual files or entire directories with recursive flag. Uses schema validation to prevent accidental deletion of critical files and enforces path safety to prevent directory traversal attacks.
Provides atomic file deletion via Obsidian's REST API with optional safety validation flags and path traversal prevention. Supports both single-file and recursive directory deletion in a single tool with schema validation.
Integrates with Obsidian's file system for consistency (unlike direct file deletion), supports optional safety checks, and prevents directory traversal attacks, whereas raw file deletion could corrupt vault state or delete critical files.
request-scoped context and observability with structured logging
Medium confidenceImplements request context tracking via RequestContext class that assigns unique IDs to each MCP tool invocation and propagates context through the entire call stack (transport layer → services → Obsidian REST API calls). Integrates structured logging with context injection, enabling correlation of logs across multiple service calls. Uses async-local-storage pattern to maintain context without explicit parameter passing. Logs all Obsidian REST API calls with request/response details and error information.
Uses async-local-storage pattern to propagate request context through the entire call stack without explicit parameter passing, enabling automatic context injection into all logs and Obsidian REST API calls. Integrates with structured logging to correlate logs across multiple service calls.
Automatic context propagation (unlike manual parameter passing) reduces boilerplate and ensures consistent context across all layers. Structured logging enables machine-readable log aggregation and correlation, whereas unstructured logs are difficult to parse and correlate.
Capabilities are decomposed by AI analysis. Each maps to specific user intents and improves with match feedback.
Related Artifactssharing capabilities
Artifacts that share capabilities with obsidian-mcp-server, ranked by overlap. Discovered automatically through the match graph.
mcp-obsidian
Model Context Protocol server for Obsidian Vaults
Obsidian MCP Server
Search, read, and write Obsidian vault notes via MCP.
Obsidian
** - Interacting with Obsidian via REST API
MCPWatch
** - A comprehensive security scanner for Model Context Protocol (MCP) servers that detects vulnerabilities and security issues in your MCP server implementations.
Webrix MCP Gateway
** - Enterprise MCP gateway with SSO, RBAC, audit trails, and token vaults for secure, centralized AI agent access control. Deploy via Helm charts on-premise or in your cloud. [webrix.ai](https://webrix.ai)
cordon-cli
The security gateway for AI agents — firewall, auditor, and remote control for MCP tool calls
Best For
- ✓AI agent developers building Claude/GPT-based knowledge management workflows
- ✓Teams integrating Obsidian as a centralized knowledge backend for LLM applications
- ✓Solo developers wanting local-first AI-assisted note management without cloud sync
- ✓LLM agents building knowledge graphs from Obsidian vaults
- ✓Developers implementing note-aware search or recommendation systems
- ✓Knowledge workers automating metadata extraction from large note collections
- ✓Security-conscious teams deploying obsidian-mcp-server with untrusted clients
- ✓Developers implementing HTTP transport exposed to the internet
Known Limitations
- ⚠Requires Obsidian Local REST API plugin running on ports 27123/27124 — no direct file system access
- ⚠HTTP transport adds network latency and requires explicit CORS configuration for cross-origin clients
- ⚠Stdio transport is single-client per process — concurrent connections require multiple server instances
- ⚠No built-in persistence or caching of vault state — relies entirely on Obsidian REST API for consistency
- ⚠Returns entire file content in memory — large notes (>10MB) may cause performance degradation
- ⚠YAML frontmatter parsing is strict — malformed YAML will cause the tool to fail rather than gracefully degrade
Requirements
Input / Output
UnfragileRank
UnfragileRank is computed from adoption signals, documentation quality, ecosystem connectivity, match graph feedback, and freshness. No artifact can pay for a higher rank.
Repository Details
Last commit: Oct 24, 2025
About
Obsidian Knowledge-Management MCP (Model Context Protocol) server that enables AI agents and development tools to interact with an Obsidian vault. It provides a comprehensive suite of tools for reading, writing, searching, and managing notes, tags, and frontmatter, acting as a bridge to the Obsidian Local REST API plugin.
Categories
Alternatives to obsidian-mcp-server
Are you the builder of obsidian-mcp-server?
Claim this artifact to get a verified badge, access match analytics, see which intents users search for, and manage your listing.
Get the weekly brief
New tools, rising stars, and what's actually worth your time. No spam.
Data Sources
Looking for something else?
Search →