hn.today

Show HN: Argus- open source security platform to protect against agentic attacks

github.com17 points4 comments

Argus is a self-hosted, MIT-licensed security operations platform designed to detect and respond to attacks across cloud services, identity systems, code, endpoints, networks, and Kubernetes. It combines telemetry collectors, detection rules, an evidence graph, and seven AI agents that investigate incidents using read-only tools. The central safety argument is that AI can help analyze threats without controlling the response: agents can suggest actions, but changes go through a separate broker with policy checks, approvals, dry runs, rollback, and a kill switch. For urgent detections, autonomous playbooks can take reversible actions from a catalog of 27 integrations, beginning in shadow mode and operating within configured limits.

The platform also includes an AI gateway for governing OpenAI and Anthropic use, discovery of unapproved AI services, and canaries that detect agents interacting with planted traps. Its testing features replay eight agentic attack scenarios and support scoped, continuous penetration testing; vulnerability findings can be enriched with CISA’s known-exploited list and EPSS scores. Audit records are hash-linked and can be mirrored to immutable storage, while backups, tenant isolation, and compliance mappings support operational controls. Docker Compose, Kubernetes Helm, and AWS and Azure Terraform options are provided. The documentation says autonomous functions are disabled by default and describes more than 460 automated tests, positioning Argus for teams that want to begin in monitoring mode and enable automation incrementally.

Read on github.com4 comments on Hacker News

Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.

More in Security

The daily digest

Today's best Hacker News stories, summarized and screenshotted, one email a day.