The Skylight team at Ai2 has built Shippy, an AI agent for real-time maritime domain awareness that prioritizes reliability through a layered architecture of skills, system prompts, and configuration. To ensure correctness when querying live data, Shippy interacts with the Skylight API via a purpose-built CLI rather than constructing raw calls, which eliminates subtle bugs related to pagination and geometry encoding.
- Shippy's capabilities are defined by versioned markdown skills that handle tasks such as querying vessel events, resolving Exclusive Economic Zone boundaries, and interpreting track data.
- The agent runs in ephemeral, isolated Kubernetes sessions provisioned by the Mothership platform, ensuring user data privacy and preventing cross-user leakage.
- Evaluation is performed using an open framework called Harbor, where subject-matter experts score agents against live data scenarios to catch regressions before deployment.
This approach allows Shippy to provide auditable, accurate decision support for maritime analysts while maintaining strict guardrails against speculation or legal determinations.