# pandanpc-inc/mcp-server [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://github.com/pandanpc-inc/mcp-server  
**GitHub Stars:** 0  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/mcp-server-94

## Description
Remote-control Claude Code, plus PandaNote notes, AI model configs & scheduled tasks.

## Tools
Capabilities this server exposes over MCP:

- **note_login** — log in to your PandaNpc account
- **sync_rules** — sync `.claude/rules/` and `.claude/skills/` Markdown to PandaNote

## Claude Desktop Quick Installation
Heuristic fallback — verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):

```json
"mcpServers": {
  "mcp-server": {
    "command": "npx",
    "args": ["-y","mcp-server-94"]
  }
}
```

## Documentation

## What pandanpc-inc/mcp-server MCP server does

The pandanpc-inc/mcp-server MCP server exposes PandaNpc functionality to MCP-compatible clients. Its tools cover several areas: PandaNote workspaces and notes, AI model and provider configuration, remote agent connections, scheduled tasks, and synchronization of local Claude Code configuration files.

For notes, the server can list, create, read, update, delete, search, and title notes. Note search supports full-text and semantic search according to the project documentation. Workspace operations include listing, creation, updates, and deletion. The server also provides tools for listing and creating models and providers, with additional model and provider operations available in the tool set.

Agent connection tools manage Claude Code, Codex, and PandaCode remote connections. Scheduled-task tools list, create, and toggle tasks, along with other scheduled-task operations. The sync_rules tool uploads Markdown from `.claude/rules/` and `.claude/skills/` to PandaNote.

## How it works

The pandanpc-inc/mcp-server MCP server is distributed through the `@pandanpc/mcp-server` npm package. Installing the package globally downloads a prebuilt binary matching the host platform. The documented binaries support Linux x64, macOS x64 and arm64, and Windows x64.

MCP clients invoke the installed `pandanpc-mcp` executable. The normal configuration uses stdio. An HTTP/SSE transport is also available by starting the executable with `--http --port 3100`. After the server starts, call `note_login` before using account-backed PandaNpc tools.

The repository contains documentation and issue tracking; the executable itself is delivered through the npm installation process rather than built from the repository source described here.

## Setup and configuration

Install the package globally with:

```bash
npm install -g @pandanpc/mcp-server
```

For Claude Code, register the command with `claude mcp add pandanpc -- pandanpc-mcp`, or add a server entry whose command is `pandanpc-mcp`. Codex uses a `pandanpc` MCP server entry in `~/.codex/config.toml` with the same command. Cursor and Windsurf can add a stdio server using `pandanpc-mcp`; they can use HTTP/SSE instead with `pandanpc-mcp --http --port 3100`.

## Tools and capabilities

The pandanpc-inc/mcp-server MCP server provides these documented tool groups:

- `note_login` for PandaNpc account authentication.
- Workspace creation, listing, updating, and deletion.
- Note CRUD operations, full-text and semantic search, and generated note titles.
- AI model and provider listing and creation, plus additional related operations.
- Agent connection listing and creation for Claude Code, Codex, and PandaCode.
- Scheduled-task listing, creation, toggling, and related operations.
- `sync_rules` for sending Claude Code rules and skills Markdown to PandaNote.

## Limitations and notes

The documentation does not specify environment variables, token names, or an alternative authentication flow; authentication is initiated through `note_login`. It also does not provide a complete enumeration of every operation represented by the ellipses in the model, provider, and scheduled-task tool groups.

The installer is described as MIT-licensed, while the distributed binary is described as proprietary freeware. Treat those as separate licensing statements when reviewing the package for commercial or redistribution requirements.

_Full upstream README: https://allmcps.com/mcp/mcp-server-94/readme_