ibm-content-services

IBM Content Services MCP Server provides a unified interface to leverage IBM FileNet Content Manager capabilities for AI workloads. The Core Server exposes essential document and content lifecycle operations such as creating, updating, checking in/out, folder management, and rich search. The Property Extraction and Classification Server extends this with automated property extraction and classification workflows, enabling AI models to derive metadata and categorize documents. The Legal Hold Server focuses on managing legal holds, allowing you to create holds and place objects under hold for compliance needs. The AI Document Insight Server (Preview) combines Content Assistant vector search with metadata filtering, enabling advanced content search, document summaries, comparisons, and Q&A capabilities by leveraging virtual tables and add-ons, assuming required components are available. To use these tools, run the respective server instances, then query the exposed endpoints or use the MCP tooling to invoke operations such as document creation, property extraction, hold management, and AI-assisted search and summarization. Each server can be used independently or together to cover end-to-end document workflows and AI-assisted analysis.

Prerequisites:

• A JavaScript runtime (Node.js) installed on the host system
• Access to IBM FileNet Content Manager environment configured for the MCP server
• If using the AI Document Insight Server: Persistent Text Extract add-on, Content Assistant add-on, and FileNet 5.7.0 IF003 or later

Installation steps:

1. Clone the MCP server repository for IBM Content Services or obtain the package distribution.
2. Install dependencies for each server module (where applicable):
   • cd core-cs-mcp-server && npm install
   • cd property-extraction-and-classification-cs-mcp-server && npm install
   • cd legal-hold-cs-mcp-server && npm install
   • cd ai-document-insight-cs-mcp-server && npm install
3. Configure each server using the provided configuration files or environment variables as described in the mcp_config section.
4. Start the servers in their preferred environment (development, staging, or production). Example: node path/to/core-cs-mcp-server/server.js node path/to/property-extraction-and-classification-cs-mcp-server/server.js node path/to/legal-hold-cs-mcp-server/server.js node path/to/ai-document-insight-cs-mcp-server/server.js
5. Verify the servers are listening on their configured ports and accessible via the MCP client tooling or API calls.

Environment variables and configuration options may vary by deployment. Ensure the Core Server is operational before enabling dependent modules like Property Extraction and Classification. The AI Document Insight Server requires several add-ons and a compatible FileNet 5.7.0 IF003+ environment. If you encounter issues with document text extraction, verify that the Persistent Text Extract add-on is installed in the object store. For search-related tools, pay attention to proper indexing and property descriptions via get_class_property_descriptions and repository_object_search. When combining servers, ensure consistent authentication and endpoint reachability across services.