SagaPeak/artifacta-mcp

What it does

SagaPeak/artifactamcp is an MCP server that functions as a centralized artifact store for AI agents. It enables persistent storage of all agent outputs, making them retrievable and shareable across multiple runs, sessions, and integrated tools. The system includes advanced features like content hash deduplication to eliminate redundant storage, expiring share links for secure sharing, and comprehensive session and agent metadata tracking.

How to set up

This tool is deployed as an MCP (Model Context Protocol) server. Installation typically involves: 1) Cloning the repository from GitHub, 2) Installing Python dependencies, 3) Configuring the MCP server settings, 4) Integrating with your AI agent or tool via MCP protocol, 5) Setting up storage backend and sharing preferences. Detailed setup instructions are available in the project's GitHub repository documentation.

Key features

• Persistent artifact storage across sessions and runs
• Content hash-based deduplication to save storage space
• Secure expiring share links for collaboration
• Session and agent metadata tracking
• 11 integrated tools for artifact management
• Path-confined uploads for security
• Destructive operations with proper safeguards
• Cross-platform compatibility (Windows, macOS, Linux)
• Cloud and local storage support