Towards a Declarative Agentic Layer for Intelligent Agents in MCP-Based Server Ecosystems

TL;DR

The paper tackles reliability issues in agentic AI by identifying a lack of explicit grounding between goals, capabilities and execution. It proposes DALIA, a declarative layer that decouples discovery, planning and execution through a Capability Semantic Model, ATDP, a federated Agent Directory MCP, and deterministic task orchestration. By grounding task graphs in declared, executable capabilities and agent permissions, DALIA enables reproducible, verifiable multi-agent workflows across heterogeneous environments. The illustrative scenario demonstrates end-to-end grounding, and the paper outlines future work toward empirical evaluation, richer capability semantics, and standardization for scalable, interoperable agentic ecosystems.

Abstract

Recent advances in Large Language Models (LLMs) have enabled the development of increasingly complex agentic and multi-agent systems capable of planning, tool use and task decomposition. However, empirical evidence shows that many of these systems suffer from fundamental reliability issues, including hallucinated actions, unexecutable plans and brittle coordination. Crucially, these failures do not stem from limitations of the underlying models themselves, but from the absence of explicit architectural structure linking goals, capabilities and execution. This paper presents a declarative, model-independent architectural layer for grounded agentic workflows that addresses this gap. The proposed layer, referred to as DALIA (Declarative Agentic Layer for Intelligent Agents), formalises executable capabilities, exposes tasks through a declarative discovery protocol, maintains a federated directory of agents and their execution resources, and constructs deterministic task graphs grounded exclusively in declared operations. By enforcing a clear separation between discovery, planning and execution, the architecture constrains agent behaviour to a verifiable operational space, reducing reliance on speculative reasoning and free-form coordination. We present the architecture and design principles of the proposed layer and illustrate its operation through a representative task-oriented scenario, demonstrating how declarative grounding enables reproducible and verifiable agentic workflows across heterogeneous environments.

Figures (3)

• Figure 1: Overview of the DALIA architecture.
• Figure 2: Example of a deterministic task graph constructed from declared capabilities and ATDP-exposed tasks.
• Figure 3: DALIA execution pipeline organised into discovery, planning and execution phases.