CAP is an open protocol that gives AI agents a structured way to perform causal reasoning — discovering variables, traversing cause-and-effect graphs, and answering "why" questions through a well-defined verb-based interface.

Most agent protocols focus on tool invocation or retrieval. CAP adds a causal layer: agents can explore causal graphs, compute interventions, and generate evidence-backed explanations — all through a transport-agnostic protocol that any runtime can implement.

• Start with cap for the CAP overview, getting started guides, and normative specification
• Use the Python SDK to build or integrate a CAP-compatible agent
• Explore cap-reference for a reference CAP server implementation

We welcome contributions! Please read our Contributing Guide and Code of Conduct before getting started.

All CAP repositories are licensed under the Apache License 2.0.