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

@snowluma/mcp

v1.10.2

Published

MCP server exposing the SnowLuma OneBot action catalog (docs + JSON Schemas) and optional execution to LLM clients.

Readme

@snowluma/mcp

An MCP server for the SnowLuma OneBot action catalog. It runs in two modes from one binary:

  • docs (default) — read-only: every action's docs, parameters, cross-field constraints, and a ready-to-use JSON Schema, so an LLM can answer "what params does set_group_ban take?" without holding the whole catalog in context.
  • execution (opt-in) — when pointed at a running OneBot HTTP endpoint, the LLM can also call actions: read-only ones freely, write ones behind a gate.

Docs only (default)

Add to your MCP client (Claude Desktop, Cline, …) — no endpoint, no execution:

{
  "mcpServers": {
    "snowluma": { "command": "npx", "args": ["-y", "@snowluma/mcp"] }
  }
}

Docs tools

  • list_actions({ category? }) — lightweight index (name / category / summary / aliases / readOnly).
  • get_action({ name }) — full doc for one action incl. inputSchema and readOnly (accepts aliases).
  • search_actions({ query }) — fuzzy match over name / summary / aliases.
  • list_categories() — categories and their action counts.

Also exposes the whole catalog as a resource: snowluma://onebot/actions.

Execution (opt-in)

Point the server at a running SnowLuma instance's OneBot HTTP endpoint (the httpServer network adapter) and it gains two execution tools:

  • query_action({ action, params? }) — calls a read-only action (e.g. get_*, can_*) and returns the full OneBot response. Annotated readOnlyHint. Refuses write actions (points you to invoke_action).
  • invoke_action({ action, params? }) — calls any known action, including ones with side effects (send a message, change a group, …). Annotated destructiveHint
    • openWorldHint. Only available in write mode.

Both pass the OneBot envelope through verbatim — a logical failure (retcode≠0) comes back as data with its wording; only a transport failure is an error.

Configuration (env)

| Variable | Meaning | | --- | --- | | SNOWLUMA_MCP_ENDPOINT | OneBot HTTP endpoint, e.g. http://127.0.0.1:3000/. Absent → docs-only (execution tools hidden). | | SNOWLUMA_MCP_TOKEN | Access token (sent as Authorization: Bearer …), if the endpoint requires one. | | SNOWLUMA_MCP_TIMEOUT_MS | Per-request timeout (default 30000). | | SNOWLUMA_MCP_MODE | docs | read | write. Default: read when an endpoint is set, else docs. |

Read mode — the LLM can query read-only actions, but cannot perform any write:

{
  "mcpServers": {
    "snowluma": {
      "command": "npx",
      "args": ["-y", "@snowluma/mcp"],
      "env": {
        "SNOWLUMA_MCP_ENDPOINT": "http://127.0.0.1:3000/",
        "SNOWLUMA_MCP_TOKEN": "your-access-token"
      }
    }
  }
}

Write mode — also enables invoke_action (the bot can send messages, manage groups, etc.). Enable deliberately:

{
  "mcpServers": {
    "snowluma": {
      "command": "npx",
      "args": ["-y", "@snowluma/mcp"],
      "env": {
        "SNOWLUMA_MCP_ENDPOINT": "http://127.0.0.1:3000/",
        "SNOWLUMA_MCP_TOKEN": "your-access-token",
        "SNOWLUMA_MCP_MODE": "write"
      }
    }
  }
}

Safety model

  • Read/write is classified per action in the source specs (by what the action actually does, not its name) and baked into the catalog. The default is write: an action is callable via query_action only if it is explicitly read-only.
  • The mode gate is enforced on every call, not just by hiding tools — calling invoke_action outside write mode is refused even if a client sends it directly.
  • Unknown actions are rejected by both tools (only catalog actions are callable), so typos and non-catalog internal actions can't be driven.
  • A well-behaved client can auto-approve query_action (read-only) and prompt for invoke_action (destructive) using the MCP tool annotations.

How it stays in sync

The catalog — including each action's readOnly flag — is a build-time snapshot generated from @snowluma/onebot's live action specs (collectActionDocs()) on every build, so it auto-tracks action add/remove and read/write reclassification. The snapshot is pinned to the SnowLuma version it was built from; a new SnowLuma release republishes a fresh catalog.

This package is generated; do not hand-edit src/generated/catalog.ts.