npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@abhishekmcp/notes

v0.4.1

Published

MCP server for local markdown notes — ranked search, tags, todos, and a wiki-link knowledge graph, from any MCP client.

Readme

@abhishekmcp/notes

An MCP server for managing local markdown notes. Lets any MCP client (Claude Desktop, Claude Code, Cursor, …) search, link, and organize the notes in a folder on your machine — with ranked full-text search, tags, todos, and a wiki-link knowledge graph.

Pure JavaScript, no native dependencies, no API keys — everything runs locally.

Features

Notes (token-efficient I/O)

  • list_notes — list notes (newest first) with pagination (offset/limit) and an optional tag filter
  • read_note — read a note; optionally just one heading's section, or a character window (offset/limit) with a truncation flag
  • get_outline — return only a note's heading tree (grasp a big note in a few tokens)
  • create_note — create a new note (optional overwrite)
  • append_note — append to a note, creating it if missing (great for journals/logs)
  • delete_note — delete a note
  • move_note — rename/move a note and rewrite every [[wiki-link]] across the vault that points at it

Search & discovery

  • search_notes — ranked full-text search (MiniSearch); supports fuzzy and prefix matching, a field filter (title/tag/body/path), and returns ranked snippets with surrounding context
  • semantic_searchmeaning-based search using local embeddings; finds related notes even with no shared keywords (e.g. "puppy" matches a note about "canine companions"). Optional hybrid mode fuses semantic + keyword ranking
  • list_tags — every tag across the vault with note counts
  • list_todos — aggregate - [ ] / - [x] checkboxes across all notes

Knowledge graph

  • get_backlinks — notes linking to a note via [[wiki-link]] syntax
  • get_neighbors — notes within N hops over the (undirected) link graph (depth/limit capped)
  • find_path — shortest wiki-link chain between two notes
  • related_notes — notes ranked by shared links + shared tags
  • graph_overview — aggregate health: note/link/tag counts, top hubs, orphans, broken-link count
  • broken_links — wiki-links that point at notes which don't exist

Organization & daily workflow

  • daily_note — open today's daily note (creating it if needed) and append a timestamped entry
  • list_templates / create_from_template — instantiate a note from a template, substituting {{date}}/{{time}}/{{title}} plus your own vars
  • rename_tag — rename a tag across the whole vault (frontmatter + inline #hashtags)
  • unlinked_mentions — find notes that mention a note's title as plain text but don't yet [[link]] to it

Prompts (slash-command workflows)

Exposed via the MCP Prompts primitive — your client surfaces these as slash commands:

  • weekly_review — summarize the last 7 days of notes + open todos
  • summarize_note — summarize one note (with note-name autocomplete)
  • daily_standup — draft a standup from yesterday/today's daily notes + open todos

Resources

  • Every note is exposed as a notes://<name> resource.

Frontmatter & tags

Notes may start with a YAML frontmatter block; title and tags (a list or comma-separated string) are recognized. Inline #hashtags in the body are also collected as tags.

---
title: My Note
tags: [project, ideas]
---
# My Note
Links to [[another-note]]. Some inline #tag too.

Configuration

All via environment variables:

| Variable | Default | Effect | |----------|---------|--------| | NOTES_DIR | ~/notes | Directory where notes live (a leading ~ is expanded). | | NOTES_READONLY | unset | Set to 1 to disable all mutating tools (create/append/delete/move are not even registered) — safe for sharing a vault. | | NOTES_NO_CACHE | unset | Set to 1 to skip the on-disk index cache and rebuild in memory each start. | | NOTES_MODEL_DIR | ~/.cache/mcp-notes/models | Where the semantic-search embedding model is cached. | | NOTES_DAILY_DIR | daily | Subdirectory (within the vault) for daily notes. | | NOTES_TEMPLATE_DIR | templates | Subdirectory (within the vault) holding note templates. |

Semantic search & the embedding model

semantic_search runs the all-MiniLM-L6-v2 model locally via WebAssembly (onnxruntime-web) — no API keys, no native dependencies, no data leaves your machine. The quantized model (~23 MB) is downloaded once on first use into NOTES_MODEL_DIR and cached; embeddings are stored in <NOTES_DIR>/.notes-embeddings.json and incrementally updated as notes change. The first semantic_search call needs network access for the download and embeds the whole vault; everything after that is offline and fast. Keyword search and all other tools work without ever triggering this.

Index cache

For fast warm starts the server persists its search index to <NOTES_DIR>/.notes-index.json and, on startup, incrementally re-parses only the notes that changed (by mtime/size) since last run. The cache is rebuilt automatically if it's missing, unreadable, or from an older index version. Files on disk are always the source of truth.

Security

All filesystem access is sandboxed to the notes directory:

  • Path traversal (../) and absolute paths are rejected.
  • Symlinks inside the vault that resolve outside it are rejected (realpath containment).
  • Single files above a size limit are refused (DoS / context guard).
  • Writes are atomic (temp file + rename), so a crash can't leave a torn note.

Usage

Claude Code — plugin (easiest)

/plugin marketplace add Abhishekkumar2021/mcp-suite
/plugin install notes

Claude Code — manual

claude mcp add notes --env NOTES_DIR=$HOME/notes -- npx -y @abhishekmcp/notes

Claude Desktop — MCPB (one-click, no Node required)

Download notes-<version>.mcpb from the latest release and drag it onto Claude Desktop → Settings → Extensions. A folder picker lets you choose your notes directory.

Claude Desktop

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "notes": {
      "command": "npx",
      "args": ["-y", "@abhishekmcp/notes"],
      "env": { "NOTES_DIR": "/absolute/path/to/your/notes" }
    }
  }
}

To share a vault read-only, add "NOTES_READONLY": "1" to env.

Develop from source

npm install                      # from the repo root
npm run build -w servers/notes
node servers/notes/dist/index.js # NOTES_DIR=... to point at a vault

Publishing to npm

Publishes automatically via GitHub Actions (Trusted Publishing / OIDC) when a release tagged notes-v<version> is created. See the repo root for the CD workflow.

License

MIT