Worlds
Worlds generates synthetic network environments — complete with hosts, services, vulnerabilities, and attack paths — that you can use for agent training and evaluation.
Core concepts
Section titled “Core concepts”Manifests
Section titled “Manifests”A manifest is a generated world graph describing a synthetic network environment. It includes hosts, services, principals, vulnerabilities, and the relationships between them. Manifests are workspace-scoped and can be inspected to understand the generated topology.
Trajectories
Section titled “Trajectories”A trajectory is a sampled attack path through a manifest. Trajectories represent sequences of actions an agent could take to navigate the environment. They can be used as training data for reinforcement learning or as evaluation benchmarks.
Trajectory sampling supports multiple modes:
- Algorithmic (
kali,c2) — deterministic sampler-based paths - Agent — agent-driven rollouts using a selected capability and runtime
Both manifest generation and trajectory sampling run as async jobs. You can track job status, cancel running jobs, and retrieve results when complete.
Project alignment
Section titled “Project alignment”Worlds resources are workspace-scoped and use project_id for grouping. If you don’t specify a project when creating a manifest, the workspace default project is used. Trajectory jobs inherit the project from their parent manifest.
Training integration
Section titled “Training integration”Worlds integrates with hosted training in two ways:
- SFT — trajectory datasets can be converted into supervised fine-tuning conversations
- RL — trajectory datasets can drive offline reinforcement learning, or manifests can be used to generate fresh agent trajectories for online RL
Artifacts
Section titled “Artifacts”Completed trajectory jobs produce training-ready artifacts:
- Training dataset (JSONL) for downstream training pipelines
- Raw trajectory records for analysis and debugging