Skip to main content
Glama
ContextKeeper MCP Server

ContextKeeper MCP Server

by ondrahracek

ContextKeeper MCP Server

npm version License: MIT

MCP (Model Context Protocol) server for ContextKeeper - integrates your context items with AI assistants like Cursor, Claude Desktop, and others.

What it does

This MCP server allows AI assistants to read and modify your ContextKeeper items directly through natural language. No need to manually switch between your editor and the terminal.

Related MCP server: AI Context Memory

Tools

Tool

Description

CLI Equivalent

list_context_items

List all context items with ID, content, project, tags, and status

ck list

add_context_item

Add a new context item

ck add

mark_context_done

Mark a context item as completed

ck done

get_context_status

Get a quick summary (counts, projects, tags)

ck status

search_context_items

Search context items by query string and/or tags

ck search

remove_context_item

Remove (archive) a context item by ID

ck remove

edit_context_item

Edit an existing context item's content and/or tags

ck edit

sync_context_items

Trigger a sync of context items to AI agent files

ck sync

Naming Convention

All tools follow MCP best practices:

  • snake_case: list_context_items, add_context_item

  • Verb-first: get_context_status, mark_context_done

  • Consistent plural: All use context_items (except status)

  • Aligned with CLI: Each tool maps directly to a ck command

Prerequisites

  • ContextKeeper CLI (ck) installed and on your PATH

  • Node.js 18 or later

  • npm or yarn

Installation

Option 1: npx (No install)

npx @ondrahracek/contextkeeper-mcp

Option 2: From Source

git clone https://github.com/ondrahracek/contextkeeper-mcp.git
cd contextkeeper-mcp
npm install
npm run build

Configuration

Cursor IDE

Add to ~/.cursor/mcp.json:

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

Or if running from source:

{
  "mcpServers": {
    "contextkeeper": {
      "command": "node",
      "args": ["/path/to/contextkeeper-mcp/build/index.js"]
    }
  }
}

Claude Desktop

Add to ~/.config/claude/claude_desktop_config.json:

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

Usage Example

In Cursor's chat, you can say:

@contextkeeper What's my current context?

Claude will respond with your active items:

[
  {
    "id": "bc2839",
    "fullId": "bc2839b5-6a8b-4b2a-9e1e-7b5c4d3e2f1a",
    "content": "change tokens - find in whatsapp with...",
    "project": "",
    "tags": ["whatsapp"],
    "completedAt": null,
    "createdAt": "2026-02-08T21:25:00Z"
  }
]

Or:

@contextkeeper Add "review security middleware"

Claude will add the item and confirm.

Development

# Install dependencies
npm install

# Run tests
npm test

# Build
npm run build

# Lint
npm run lint

Architecture

contextkeeper-mcp/
├── src/
│   ├── index.ts       # MCP server entry point
│   ├── helpers.ts     # CLI wrapper functions
│   └── types.ts       # TypeScript type definitions
├── build/             # Compiled JavaScript output
└── package.json

License

MIT - see LICENSE for details.

Contributing

Contributions welcome! Please open issues or PRs on GitHub.

Available Tools

9 tools
add_context_itemA

Add a new context item to ContextKeeper. Use this to track tasks, ideas, or notes.

Inputs: content (required), path, sync, tags, +1 more.

edit_context_itemB

Edit an existing context item's content and/or tags using its ID.

Inputs: id (required), path, sync, tags, +1 more.

get_context_statusA

Get a summary of the context: total items, completed vs active counts, projects, and tags.

Inputs: path.

init_contextA

Initialize a new ContextKeeper directory with .contextkeeper folder and items.json file.

Inputs: path.

list_context_itemsB

List all context items from ContextKeeper. Shows ID, content, project, tags, and completion status.

Inputs: path, showCompleted.

mark_context_doneA

Mark a context item as completed using its ID (full UUID or 6-char prefix).

Inputs: id (required), path.

remove_context_itemA

Remove (archive) a context item by ID. The item is soft-deleted and can be recovered.

Inputs: id (required), path, sync.

search_context_itemsA

Search context items by query string and/or tags. Returns matching items with their details.

Inputs: path, tags, query, showCompleted.

sync_context_itemsB

Trigger a sync of context items to AI agent files. Updates the .context-keeper directory.

Inputs: path.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.8/5.0

    Scored across 9 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: list vs search vs status, CRUD operations (add, edit, mark_done, remove), and setup/sync (init, sync). No two tools could be easily confused.

    Naming Consistency5/5

    All tools follow a consistent snake_case verb_noun pattern: list_context_items, add_context_item, mark_context_done, get_context_status, search_context_items, remove_context_item, edit_context_item, sync_context_items, init_context.

    Tool Count5/5

    9 tools is well-scoped for a context management server, covering all essential operations without redundancy or bloat.

    Completeness4/5

    Core CRUD and lifecycle are covered (create, read, update, delete, status, search, sync). Minor gap: soft-deleted items are described as recoverable but no explicit recovery tool is exposed, which is a small dead end.

    Maintenance

    ActivityInactive
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      Not graded
      maintenance
      An MCP server enabling AI assistants to store, retrieve, and manage contextual information across conversations with features like persistent memory, advanced search, tagging, and privacy controls.
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      An MCP server that enables AI assistants to manage Context Repo prompts, documents, and collections with semantic search and progressive disclosure navigation.
      25 npm
      1
      MIT