margins-mcp
Related Servers
Alternatives to margins-mcp
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityDmaintenanceConnects Claude Desktop to your Obsidian vault, enabling reading, writing, searching, and organizing notes locally.1MIT
- AlicenseAqualityDmaintenanceConnects Claude Desktop to an Obsidian vault, enabling reading, searching, capturing ideas, and managing notes through natural language.213 npmMIT
- AlicenseNot gradedqualityBmaintenanceConnects Claude Desktop and Claude Code to an Obsidian vault by exposing notes as tools, enabling folder listing, note reading, text and semantic search, tag and link queries, and optional note creation, editing, and management, all locally without installing Node.js or editing JSON.1MIT
- AlicenseNot gradedqualityDmaintenanceConnects Claude.ai to your local Obsidian vault for full CRUD access, search, and daily note creation via the Model Context Protocol.67 npm14MIT
- AlicenseNot gradedqualityBmaintenanceEnables Claude.ai and Claude Code to read and write an Obsidian vault from anywhere over OAuth-protected HTTP, including notes, properties, tags, links, canvases, daily notes, and attachments, without Obsidian running or the vault being on the local machine.MIT
- AlicenseNot gradedqualityCmaintenanceConnects Claude to a local Obsidian vault, enabling listing, reading, searching, and optionally editing notes via natural language, with all operations running locally.3,512 npmMIT
TDQS
Scored across 23 tools
Most tools are clearly separated by verb_noun intent, but there are confusing pairs: search_vault vs. search, and read_page vs. fetch, which both return text for a vault path. The detailed descriptions help, but an agent could easily misselect between these overlapping read/search tools.
The dominant pattern is verb_noun, with list_, get_, propose_, and record_ tools carrying most of the surface. However, the mix is inconsistent: three tools use a margins_ prefix, two bare tools are just search and fetch, and recall_preferences breaks the get/list convention.
23 tools is at the heavy end for a single MCP server and feels like several subsystems (vault context, raw ingestion, proposal management, research, diagnostics) bundled together. Each tool has a real use in the workflow, but the set would be easier to reason about if split or trimmed.
The core lifecycle is well covered: ingest raw files, propose compilations, edit or append to pages, list/resolve proposals, and read/search the vault. Obvious gaps include no explicit delete/remove tool for vault pages and no rename/move operation, but these are workarable rather than workflow-breaking.