ContextVault MCP

Overview

ContextVault provides a shared memory layer for AI and teams, featuring a native MCP server for cross-client access across ChatGPT, Codex, Claude, and Gemini. It enables persistent, shared context that AI agents can access and update across different sessions and tools.

Features

• Shared memory layer: Persistent context accessible across AI clients
• Cross-client compatibility: Works with ChatGPT, Codex, Claude, and Gemini
• Native MCP server: Standard MCP integration for AI tools
• Team collaboration: Shared context for team-based AI workflows
• Session persistence: Context survives across sessions and restarts

Installation

npm install -g @contextvault/mcp

Configuration

{
  "mcpServers": {
    "context-vault": {
      "command": "contextvault",
      "args": ["mcp"],
      "env": {
        "CONTEXTVAULT_API_KEY": "your-api-key"
      }
    }
  }
}

Available Tools

Tool	Description
get_context	Retrieve shared context by key or topic
set_context	Store or update shared context
list_context	List available context entries
delete_context	Remove context entries

Usage Examples

# Initialize ContextVault
contextvault init --name my-team-context

# Start MCP server
contextvault mcp

# Access context from any AI client

Pros

• ✅ Shared context across multiple AI clients
• ✅ Persistent memory for AI agents
• ✅ Team collaboration features
• ✅ Native MCP integration

Cons

• ❌ Requires cloud service or self-hosted instance
• ❌ May have latency for context lookups
• ❌ New project, limited track record

When to Use

• You need shared context across multiple AI tools
• You're building team-based AI workflows
• You want persistent memory for AI agents across sessions

Resources

• GitHub
• Documentation