The Actian MCP Server for Analytics Engine connects AI applications to the Actian Analytics Engine database using the open Model Context Protocol (MCP). It acts as a secure bridge between MCP-compatible AI clients and your database, enabling agents to discover schemas, execute queries, and retrieve real-time context through a standardized interface.
The Analytics Engine is suited for large-scale retrieval-augmented generation (RAG) and high-performance vector analysis.
| Action | Description |
|---|---|
| Run SQL queries | Read-only by default; enable read-write mode to also run INSERT, UPDATE, DELETE, and MERGE. Data Definition Language (DDL) is always rejected in both modes. |
| Human-in-the-loop write approval | Every write requires a person to approve it in the connected MCP client, and the OAuth mcp:write scope when OAuth is enabled. |
| List tables and views | Discover available objects in the schema. |
| Inspect table structure | Retrieve column definitions, types, and comments. |
| Read schema metadata | Explore database-level metadata. |
| List functions and procedures | View available user-defined routines. |
| Custom extensions | Add custom tools and resources with Python modules mounted at /app/extensions, without modifying the core plugin. |
docker run -d \
-p 8000:8000 \
-v ./conf.json:/app/conf.json \
actian/analytics-engine-mcp-server:1.1.0
For configuration file format, HTTPS setup, OAuth authentication, and MCP client integration, see the full documentation.
The Actian MCP Server software contained in this image is copyright (c) 2026 Actian Corporation. See /app/ACTIAN_EULA.txt in the image for license terms.
Content type
Image
Digest
sha256:44e7f8220…
Size
123.7 MB
Last updated
about 2 months ago
docker pull actian/analytics-engine-mcp-server:1.1.0