Agent Settlement Extension (ASE) is an economic metadata layer that extends existing agent-to-agent (A2A) and Model Control Protocol (MCP) communication protocols with economic semantics. ASE provides standardized schemas, validation, and reference implementations to enable agents to express economic intents, settlements, audit bundles, and related metadata in interoperable ways.

The key goals of ASE include making economic semantics first-class in agent messaging, providing machine-readable schemas and validators for settlement, audit, and delegation tokens, and offering lightweight reference code to integrate ASE with agent frameworks. The system includes standardized structures for cost attribution, delegation, and settlement events, enabling agents to exchange not just tasks, but economic intent.

ASE works by providing JSON Schema files describing ASE data structures including audit bundles, delegation tokens, monetary amounts, and other economic metadata. The reference implementation includes core models, validation, serialization, business logic, integration adapters for frameworks like LangChain and AutoGPT, key handling and signing utilities, and compliance helpers with RFC-style governance workflows.

The benefits include enabling governed, auditable agent economies and providing interoperability between ASE-aware and non-ASE agents. ASE is designed to be framework-agnostic, with adapters demonstrating integration patterns with popular agent frameworks.

ASE targets developers working with agent frameworks who need to add economic semantics to their agent communications. It includes Python reference implementations and is licensed under Apache License 2.0, making it suitable for open source and commercial projects requiring economic metadata capabilities.