AI agent integrations
Use ARMO directly from your AI assistant — list incidents, triage CVEs, manage exception policies, and more, in natural language.
An AI agent integration connects your AI assistant directly to the ARMO security platform. Once installed, the agent can read findings across all 13 resource clusters — incidents, vulnerabilities, posture controls, risks, attack chains, inventory, network policies, seccomp profiles, runtime rules, runtime policies, integrations, cloud accounts, and repository posture — and perform safe mutations such as creating exceptions, generating policies, or opening Jira tickets. Every mutating action goes through a dry-run preview before it executes, so the agent can show you exactly what will change before committing.
The plugin or extension format is platform-specific (Claude Code uses a plugin manifest; Gemini CLI uses an extension manifest), but the underlying tool is the same: armoctl, a JSON-first CLI built for both human and agent use. Install armoctl once, configure it with your ARMO credentials, and it works identically whether called by you at a terminal or invoked by an AI assistant responding to a natural-language request.
Supported agents:
More agent integrations and an MCP server are planned. Watch the ARMO changelog for announcements.
Updated about 2 hours ago