R-LAM is a reproducibility-constrained execution framework for Large Action Models in scientific workflow automation. It enables adaptive, agent-driven workflow execution while enforcing strict guarantees on auditability, determinism, and replayability.

OpenExecution Provenance Specification — implements AEGIS (Agent Execution Governance and Integrity Standard) for auditable, tamper-evident AI agent behavioral records. Apache 2.0.

Execution provenance protocol for AI agents — tamper-evident, third-party verifiable behavioral records. Hash chains (SHA-256) + Ed25519 signatures + JCS (RFC 8785). Zero dependencies, 202 tests.