Vectra AI RUX MCP Server — Claude Code MCP server setup
Vectra AI RUX MCP Server brings Vectra AI security platform telemetry and incident workflows directly into your AI assistant via MCP tools.
no license updated 2026-10-29
Open on GitHubWhat Vectra AI RUX MCP Server does
The server exposes Vectra AI platform capabilities to AI models through standard tool calls. Agents can inspect detection telemetry, correlate security events using plain prompts, compile investigation reports, and trigger incident response actions without leaving the chat interface. It also pulls health metrics across platform subsystems, including EDR connectors and Network Brain.
Running the server requires the uv tool. Isolated execution runs through uvx vectra-ai-mcp-server, automatically pulling Python 3.12+ when needed. Single-tenant deployments pull credentials from VECTRA_BASE_URL, VECTRA_CLIENT_ID, and VECTRA_CLIENT_SECRET.
Multi-tenant environments read a tenants.yaml configuration file. In this setup, tools receive tenant-specific prefixes such as prod_ or staging_, while individual API credentials can be injected as VECTRA_TENANT_<NAME>_<FIELD> variables.
What it helps with
- query threat detection telemetry in natural language
- trigger incident response actions inside Vectra
- inspect Network Brain and EDR platform health
- generate security investigation summaries
Example prompts for the agent
- Check the health status of external connectors and Network Brain
- Summarize recent threat detections and draft an incident report
- List available tenants and show active staging alerts
How to install Vectra AI RUX MCP Server
Install
The MCP registry has no launch command for it: see the repository README
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Vectra AI RUX MCP Server
How to install Vectra AI RUX MCP Server?
Install uv on your system and add vectra-ai-mcp to your MCP client config using command: "uvx", args: ["vectra-ai-mcp-server"], and specify your portal URL alongside API credentials in the env section.
What transports does the server support?
The server provides standard stdio transport for desktop MCP clients, along with sse and streamable-http modes for HTTP network connections.
How does multi-tenant configuration work?
Supply a path with --config /path/to/tenants.yaml. Tools get isolated per tenant with prefixes like prod_list_detections, supplemented by a list_tenants discovery tool.