The 1c MCP Toolkit by Roctup enables AI agents to interact with 1C:Enterprise databases. This free solution operates as an MCP server, primarily using 1C Enterprise script for its built-in server mode. A proxy mode is also available, which incorporates Python and supports Docker. The toolkit is distributed as an external processing file for 1C:Enterprise or via Python scripts. It provides essential capabilities for integrating modern AI systems with established 1C environments, bridging data and processing gaps.

What does 1c MCP Toolkit connect to?

The 1c MCP Toolkit connects directly to 1C:Enterprise databases, functioning as an MCP server. Its primary goal is to bridge AI agents with these established database systems. The toolkit is predominantly built with 1C Enterprise script (BSL) for its recommended "built-in server" mode, which operates without requiring Python. For situations demanding a "proxy mode," Python is utilized, making the solution adaptable to diverse deployment strategies within the 1C ecosystem. The project's language breakdown highlights 1C Enterprise (47.9%) and C++ (38.5%).

Core features and tools

The toolkit supports HTTP as a transport protocol, available for both its built-in server and proxy modes. It additionally provides a REST API, serving as an alternative to the Model Context Protocol. Confirmed compatible AI clients include Kiro and Claude Desktop. For Claude Desktop users, connecting to HTTP servers may necessitate an intermediary like mcp-remote, which itself requires Node.js. The toolkit mentions "legacy SSE" in its authentication details, indicating a breadth of protocol considerations.

Who is it for?

The 1c MCP Toolkit is designed for 1C developers and users who aim to integrate AI agents directly with their 1C:Enterprise databases. It enables AI agents, such as Kiro and Claude, to interact with 1C systems for essential data retrieval and processing tasks. This makes it a valuable asset for environments where automated access to 1C data is a priority for AI-driven applications and workflows. Skip this if your project does not involve 1C:Enterprise databases or AI agent integration.

How to get started

The built-in server authenticates using a token-based model. Setup involves providing an "Authorization: Bearer YOUR_TOKEN" header when connecting via HTTP.

Use cases

• Obtaining metadata from 1C databases.
• Accessing and querying data within 1C databases.
• Enabling AI agents, including Kiro and Claude, to interact with 1C systems for data retrieval and processing.

The 1c MCP Toolkit offers a direct path for AI agents to engage with 1C:Enterprise databases. It supports both a built-in server mode and a proxy mode, accommodating different deployment needs. With HTTP transport and a REST API, it provides flexibility for integration. The token-based authentication model ensures secure access.