Graph-native MCP server for Obsidian — search, navigate, and safely edit your vault from Codex, ChatGPT, Claude, and any MCP client.
claude mcp add obsidian-everywhere -- npx -y obsidian-everywhere{
"mcpServers": {
"obsidian-everywhere": {
"command": "npx",
"args": ["-y", "obsidian-everywhere"]
}
}
}Resumen de MCP Servers
<div align="center">
[English](README.md) | [한국어](README.ko.md)
# 🧠 Obsidian Everywhere
**Your Obsidian vault, as a graph, in Codex, ChatGPT, Claude, and other MCP clients.**
[](https://github.com/junnnnnw00/obsidian-everywhere/actions/workflows/ci.yml)
[](LICENSE)
[](package.json)
[](tsconfig.json)
[](https://modelcontextprotocol.io)
[](https://www.npmjs.com/package/obsidian-everywhere)
[](CONTRIBUTING.md)
*Codex CLI · ChatGPT Desktop (Codex) · Claude Code/Desktop · remote clients — one server, every surface.*
</div>

---
**This is a graph server, not a markdown file server.** Your AI client shouldn't see
your vault as "a folder of `.md` files" — it should see notes and links as a
graph: backlink traversal, n-hop neighborhoods, and topic-centered context
bundles are first-class tools, not an afterthought bolted onto a file
reader. Unresolved links stay in the graph (that's a real signal about your
vault, same as it is in Obsidian itself), and every response is structured
for what an LLM actually needs — explicit link relationships, not just raw
text.
## Contents
- [Features](#features)
- [Try it without your vault](#try-it-without-your-vault)
- [Why Obsidian Everywhere?](#why-obsidian-everywhere)
- [Where does this actually run?](#where-does-this-actually-run)
- [Quickstart](#quickstart)
- [Configuration](#configuration)
- [Development](#development)
- [Project status](#project-status)
- [Contributing](#contributing)
- [License](#license)
## Features
```
vault (.md files)
│ parse · watch
▼
SQLite index (FTS5) ⇄ in-memory graph (graphology)
│ n-hop · shortest path · PageRank
▼
31 MCP tools
│
▼
stdio · bearer-token HTTP · OAuth HTTP
```
- 🧩 **Real graph engine** — a markdown parser (wikilinks, embeds,
frontmatter, nested tags, headings, block references), a SQLite index
with full-text search, and an in-memory [graphology](https://graphology.github.io/)
layer for n-hop traversal, shortest paths, and PageRank — kept in sync
incrementally as files change, never rebuilt from scratch.
- 🛠️ **31 graph-native MCP tools** — graph navigation, structured/paginated reads, safe lifecycle and partial edits, rollback-capable bulk cleanup, regex/listing, Base checks, and persisted Obsidian settings.
- 🔌 **Three ways to connect** — stdio for local MCP clients (including
Codex CLI, ChatGPT Desktop, and Claude), Streamable HTTP with a static
bearer token for private remote clients, and Streamable HTTP with OAuth
2.1 (PKCE + Dynamic Client Registration) for public connectors.
<details>
<summary><strong>Full tool list</strong></summary>
**Read**
| Tool | What it does |
|---|---|
| `vault_overview` | Note counts, top tags, PageRank hub notes, recently modified — a starting orientation |
| `search_notes` | Full-text search with tag/folder filters, each result annotated with link counts and tags |
| `read_note` | Structured content/frontmatter/links/tags plus line pagination; optional heading-scoped read |
| `list_notes` | Explicit folder-aware note listing with pagination |
| `list_folder` | Immediate child folders, notes, and attachments |
| `regex_search` | JavaScript-regex search with file, line, and excerpt |
| `get_backlinks` | Every note linking to a given note, with the linking sentence |
| `get_neighborhood` | Explicit n-hop node/edge list around a note (links treated as undirected) |
| `get_context_bundle` | **The killer feature.** Center note + prioritized 1-hop neighbors packed into a token budget |
| `list_tags` | Full nested tag hierarchy with counts |
| `get_notes_by_tag` | Notes carrying a given tag (nested-aware) |
| `find_orphans` | Notes with no incoming or outgoing links |
| `find_unresolved` | Links that don't resolve to any note, grouped by target |
| `find_path` | Shortest connection path between two notes, with a one-line summary per hop |
| `get_related` | Similar notes that *aren't* directly linked yet (Jaccard similarity over shared tags/neighbors) |
| `get_hotkeys` / `get_obsidian_settings` | Persisted hotkey command IDs, Templates folder, and core-plugin settings |
| `validate_base` | Static YAML/shape validation for `.base` files or fenced Base blocks |
**Write**
| Tool | What it does |
|---|---|
| `create_note` | Create a new note (with frontmatter); reindexed immediately — the next tool call already sees it |
| `append_to_note` | Append to a note, optionally under a specific heading; fails closed if the heading isn't found |
| `move_note` / `rename_note` / `delete_note` | Lifecycle operations with inbound-link rewriting, backlink guardrails, and recoverable trash |
| `replace_text` / `patch_section` | Guarded exact-text and heading-scoped edits |
| `update_frontmatter` / `remove_frontmatter_field` | Change properties without replacing the note body |
| `bulk_replace` / `rollback_bulk_edit` | Dry-run-first folder/regex replacement with snapshots and rollback |
| `set_hotkey` / `set_templates_folder` | Update persisted Obsidian settings (vault reload may be required) |
Write tools are on by default for stdio and the
bearer-token HTTP transport, and off by default for the public OAuth
connector transport (opt in with `OAUTH_ENABLE_WRITE_TOOLS=true`) — see
[Configuration](#configuration) and DECISIONS.md D15.
</details>
## Try it without your vault
Run the built-in demo first. It creates a temporary sample vault, shows graph
orientation and unresolved-link discovery, previews a safe bulk edit, and then
removes the sample. It never reads or changes your own notes.
```bash
npx -y obsidian-everywhere demo
```

When you are ready to connect a real vault, generate copyable configuration for
Codex, ChatGPT Desktop, Claude Code, and Claude Desktop:
```bash
npx -y obsidian-everywhere init /absolute/path/to/your/vault
npx -y obsidian-everywhere doctor /absolute/path/to/your/vault
```
`init` only prints configuration—it never edits global client settings.
`doctor` checks Node.js, permissions, Obsidian metadata, SQLite, parsing, and the
graph engine without printing note content. Add `--share` to redact the vault
path before pasting diagnostics into an issue.
## Why Obsidian Everywhere?
There are several good Obsidian MCPs. Pick the architecture that matches how
you work rather than assuming one server wins every category.
| | **Obsidian Everywhere** | [obsidian-mcp-server](https://github.com/cyanheads/obsidian-mcp-server) | [Local REST API](https://github.com/coddingtonbear/obsidian-local-rest-api) | [TurboVault](https://github.com/epistates/turbovault) |
|---|---|---|---|---|
| Install | `npx` | `npx` | Obsidian community plugin | `cargo install` / binary |
| Published tools | **31** | 14 | 16 | 74 |
| Obsidian must be open | **No** | Yes | Yes | **No** |
| Best graph capability | PageRank, shortest path, n-hop, unresolved links | Outgoing links in structured reads | Live Obsidian metadata/search | Multi-hop, centrality, clusters, suggestions |
| Safe editing | Partial edits; bulk dry-run, snapshot, rollback | Surgical edits and frontmatter/tag management | Live heading/block/frontmatter patching | Conflict hashes, audit rollback, Git-backed batch |
| Live app commands/current file | Persisted settings only | **Yes** | **Yes** | No |
| Remote transport | stdio, bearer HTTP, **OAuth 2.1** | stdio, HTTP with JWT/OAuth | HTTP with API key | stdio, HTTP, WebSocket, TCP |
| Best fit | Fast `npx`, focused LLM context, graph navigation, safe vault cleanup | Rich app-driven CRUD and Omnisearch | Direct control of a running Obsidian app | Maximum breadth, multi-vault and advanced analysis |
Comparison checked against each project's published documentation on
2026-07-20. A blank or narrower cell means “not documented there,” not that a
project can never support it. If you need active-file state or command-palette
execution, choose a plugin-backed server. If you want a headless, one-command
graph server with token-budgeted context and guarded cleanup, that is the niche
Obsidian Everywhere is designed for.
Everything runs locally by default. There is no account, API key, hosted vault,
or telemetry requirement.
See [`docs/architecture.md`](docs/architecture.md) for how it's built and
[`docs/deploy.md`](docs/deploy.md) for the full deployment topology
(LaunchAgent, Docker, Cloudflare Tunnel).
## Where does this actually run?
**The `obsidian-everywhere` process needs direct filesystem access to your
vault's `.md` files** (to parse them, watch for changes, etc.) — so it
must always run on **the machine where your vault physically lives**
("the vault machine": your laptop, most likely). It does not matter which
client machine you're working from — the *server* always runs on the vault
machine; only the *client* connection method changes.
| Where you use the MCP client | What you need |
|---|---|
| The same machine as the vault | **stdio.** Nothing else — Codex, ChatGPT Desktop, Claude Code/Desktop, or another local client spawns the server directly. |
| A different machine you control (a lab/work server, another laptop, an SSH box) | **Bearer-token HTTP** + a private network between the two machines (we recommend [Tailscale](https://tailscale.com/dowLo que la gente pregunta sobre obsidian-everywhere
¿Qué es junnnnnw00/obsidian-everywhere?
+
junnnnnw00/obsidian-everywhere es mcp servers para el ecosistema de Claude AI. Graph-native MCP server for Obsidian — search, navigate, and safely edit your vault from Codex, ChatGPT, Claude, and any MCP client. Tiene 2 estrellas en GitHub y se actualizó por última vez today.
¿Cómo se instala obsidian-everywhere?
+
Puedes instalar obsidian-everywhere clonando el repositorio (https://github.com/junnnnnw00/obsidian-everywhere) o siguiendo las instrucciones del README en GitHub. ClaudeWave también te ofrece bloques de instalación rápida en esta misma página.
¿Es seguro usar junnnnnw00/obsidian-everywhere?
+
junnnnnw00/obsidian-everywhere aún no ha sido auditado por nuestro agente de seguridad. Revisa el repositorio original en GitHub antes de usarlo en producción.
¿Quién mantiene junnnnnw00/obsidian-everywhere?
+
junnnnnw00/obsidian-everywhere es mantenido por junnnnnw00. La última actividad registrada en GitHub es de today, con 9 issues abiertos.
¿Hay alternativas a obsidian-everywhere?
+
Sí. En ClaudeWave puedes explorar mcp servers similares en /categories/mcp, ordenados por popularidad o actividad reciente.
Despliega obsidian-everywhere en tu cloud
Lleva este repo a producción en minutos. Cada plataforma genera su propio entorno con variables de entorno editables.
¿Mantienes este repo? Añade un badge a tu README
Pega el badge en tu README de GitHub para mostrar que está auditado por ClaudeWave. Cada badge enlaza de vuelta a esta página y muestra el Trust Score actual.
[](https://claudewave.com/repo/junnnnnw00-obsidian-everywhere)<a href="https://claudewave.com/repo/junnnnnw00-obsidian-everywhere"><img src="https://claudewave.com/api/badge/junnnnnw00-obsidian-everywhere" alt="Featured on ClaudeWave: junnnnnw00/obsidian-everywhere" width="320" height="64" /></a>Más MCP Servers
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
An open-source AI agent that brings the power of Gemini directly into your terminal.
The fastest path to AI-powered full stack observability, even for lean teams.
Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!