Full course coming soon

Runtime Security for Autonomous Systems

7 chapters · 31 lessons

1. 1. Threat Modeling for Autonomous Actors

   4 lessons

   Identify risks introduced when non-human agents make independent decisions about file access, command execution, and tool invocation.

2. 2. Policy Enforcement Architecture

   5 lessons

   Build runtime policy checks that evaluate agent actions against allowlists, scopes, and risk thresholds before execution.

3. 3. Least Privilege and Credential Scoping

   4 lessons

   Apply zero-trust principles to limit agent access to only the resources and permissions required for each task.

4. 4. Sandboxing and Execution Isolation

   4 lessons

   Contain agent actions within isolated environments to prevent lateral movement and limit blast radius.

5. 5. Telemetry and Audit Logging

   5 lessons

   Capture structured logs of agent behavior to support investigation, compliance, and anomaly detection.

6. 6. Anomaly Detection and Behavioral Monitoring

   4 lessons

   Detect suspicious agent behavior by comparing runtime actions to expected patterns and risk baselines.

7. 7. Operationalizing Agent Security in Teams

   5 lessons

   Integrate agent security controls into developer workflows, CI/CD pipelines, and incident response processes.