Skip to content

Anthropic

Anthropic provides the AI models that power our agent capabilities.

Why Anthropic

Claude offers the best combination of capability and safety for autonomous agent workloads:

  • Tool use: Native support for function calling and tool use
  • Long context: Extended context windows for complex tasks
  • Safety: Constitutional AI approach aligns with our capability-scoped execution model
  • Reliability: Consistent, high-quality outputs

How We Use It

Claude powers the AI reasoning in our platform:

  • SmartBoxes: Claude as the coding agent within sandboxed environments
  • Murphy: Claude for delivery analysis and intervention recommendations
  • P4gent: Claude for long-running autonomous workflows

Provides Primitives