You want AI that actually does things — not a chatbot that summarizes your docs, but agents that talk to your infrastructure, run your tools, and come back with answers. We built the orchestration layer for that.
Prospector Studio
The AI platform. This is where agents live, knowledge bases get built, and workflows run. Think of it as the brains behind everything else on the platform.
- Build agents — configure LLM-backed agents with custom tool access, system prompts, and domain-specific knowledge
- Knowledge bases — RAG-powered document stores that give your agents context about your environment, policies, and procedures
- Visual workflows — drag-and-drop automation that chains agent actions, tool calls, and decision logic
- MCP servers — connect external tools and data sources using the Model Context Protocol
- Chat interface — talk to your agents directly, or let them work autonomously in the background
Get started with Prospector Studio →
Read the docs →
GraphQL API
Programmatic access to the whole platform. When you want to integrate Prospector Studio into your own tooling, scripts, or CI/CD pipelines.
- Full API coverage — everything you can do in the UI, you can do through the API
- Subscriptions — real-time updates via GraphQL subscriptions, so you can react to events as they happen
- Trigger workflows — kick off automation from external systems
- Query agents — ask questions and get structured responses programmatically
How these fit into the platform
Prospector Studio is the platform backend — it's the hub that all connectors register with. The agents here can reach into any connected tool (KubeStudio, Pick, StrikeKit, and anything you build with the SDK). For the full picture, see How It All Fits Together.