Agentdiscover is an open-source security tool that discovers every AI agent in an organization's infrastructure. It performs multi-layer detection including static code analysis, network monitoring, eBPF and Kubernetes runtime observation, endpoint inspection, and CloudTrail log auditing. It supports company-level correlation, generates AIBOM (AI Bill of Materials) reports, and detects MCP servers.
agentdiscover is a LLM eval & observability product. It focuses on identifying and governing unknown AI agents running inside corporate infrastructure. agentdiscover is an open-source project aimed at security and platform engineers. The project is open source (MIT). agentdiscover is available on the command line, and it can be self-hosted.
Behind agentdiscover is Defend AI Tech Inc., and the product first shipped in 2025. The project is developed in the open on GitHub with 18 stars and 21 commits in the last 90 days. Among its 7 catalogued features are 5-Layer Detection, Static Analysis, and eBPF Monitoring. It exposes integrations via an MCP server.
Latest indexed changes and source events
agentdiscover verified by the PulseGate indexer
Other apps tracked under the same category.
