🤖 AI Summary
This work addresses the lack of a unified configuration governance mechanism in heterogeneous multi-agent systems, which hinders versioned, auditable, and cross-framework consistent management. The authors propose a framework-agnostic reference model for agent configuration governance that normalizes diverse configurations into a canonical configuration graph via semantic projection and enforces uniform governance semantics over this graph. Key innovations include typed and independently versioned configuration items, strict decoupling of configuration from runtime, a lattice-based monotonic influence propagation mechanism, and dependency-aware immutable revisions with provenance tracking. The model is validated across LangGraph, CrewAI, and OpenAI Agents SDK, demonstrating governance-equivalent ACM representations across 27 governance scenarios and 9 propagation cases, while guaranteeing convergence, termination, and a unique fixed point.
📝 Abstract
Agentic systems are increasingly composed of heterogeneous agents, prompts, tools, models, skills, composite subsystems, policies, and execution workflows whose configurations evolve across frameworks and runtime environments. Existing LLMOps and AgentOps platforms support orchestration and observability but do not provide a common configuration-governance model for representing and governing these systems as coherent, versioned configurations.
This paper introduces Agentic Configuration Management (ACM), a framework-independent governance and configuration reference model for heterogeneous agentic systems. ACM combines typed and independently versioned Agentic Configuration Items, immutable revisions and baselines, explicit configuration-runtime separation, lifecycle and assurance semantics, dependency-aware impact propagation, and runtime provenance. Heterogeneous native configurations are normalized through semantic projection into a canonical Configuration Graph on which common governance semantics operate.
We provide a Python reference implementation with adapters for LangGraph, CrewAI, and the OpenAI Agents SDK. The evaluation combines 27 governance scenarios with nine quantitative impact-propagation cases. For the evaluated configurations, the three frameworks yield governance-equivalent ACM representations and reproducible governance outcomes after projection. The impact semantics are formalized as monotone propagation over a finite lattice, establishing convergence, termination, and uniqueness of the least fixed point above the initial impact valuation. These results provide evidence that common governance semantics can support reproducibility, auditability, dependency analysis, and interoperability across heterogeneous agentic execution abstractions within the evaluated scope.