Step 1 5 min

What Is Agent Discovery?

Understand how MOSS scans your infrastructure to find AI agents you did not register.

The Shadow AI Problem

AI agents are running in your org right now that you do not know about.

Developers deploy AI agents quickly: a Lambda function with an OpenAI call, a Kubernetes pod running a Llama model, a CI/CD pipeline that uses GPT-4 for code review. These are all AI agents making decisions that affect your business. If they are not registered with MOSS, they are shadow AI: untracked, ungoverned, and invisible to your compliance team.

MOSS Agent Discovery scans your infrastructure using 10 different scanners to find every AI agent, classify its sensitivity, and give you a one-click option to register it.

The 10 Scanners

AWS Scanner

Bedrock agents, Lambda with AI SDKs, SageMaker endpoints

Azure Scanner

Azure OpenAI deployments, AI services, Bot Framework

GCP Scanner

Vertex AI agents, Cloud Run AI services, Dialogflow

Kubernetes Scanner

AI workloads in pods, deployments, and jobs

Docker Scanner

AI containers running locally or in production

CI/CD Scanner

AI agents in GitHub Actions, GitLab CI, Jenkins

OpenTelemetry Scanner

Discovers agents from OTel traces and spans

SIEM Scanner

Discovers agents from SIEM logs (Splunk, Elastic)

SDK Footprint Scanner

Finds MOSS SDK installations in your code repos

Behavior Signature Scanner

Infers agent behavior from observed patterns

How It Works

Step 1

Connect scanners

Step 2

Run scan

Step 3

Review discovered

Step 4

Auto-register