aiqbench analysis
This research demonstrates that while declarative skill files enhance procedural accuracy and reduce orchestration errors in well-retrieved environments, they cannot compensate for poor retrieval quality, which remains the dominant bottleneck for all agent architectures. The findings suggest that improving retrieval mechanisms is a prerequisite for realizing the benefits of flexible, natural-language-based agent control flows. Consequently, practitioners should prioritize data quality and retrieval robustness before relying on complex skill definitions to mitigate agent brittleness.
Declarative Skills for AI Agents in Knowledge-Grounded Tool-Use Workflows
arXiv:2606.06923v1 Announce Type: new Abstract: We study orchestration mechanisms for tool-using AI agents in realistic customer-service workflows over an unstructured knowledge base. We argue that declarative agents -- AI agents equipped with natural-language skill files appended to the system prompt -- are an effective orchestration paradigm. Concretely, we compare (i) a DeclarativeAgent that reads three domain-specific skill files at inference time and decides its own control flow, (ii) an ImperativeAgent based on a programmatic state machine with explicit phases, and (iii) an unscaffolded baseline agent modeled after the $\tau$-Knowledge benchmark agent. Our ImperativeAgent is motivated by externalised-control inference as in Recursive Language Models and graph-based orchestration frameworks. We formalise the three agents as policy classes within a decentralised partially-observable Markov decision process and analyse their information-theoretic and structural properties; we then test the predicted differences empirically on five language models and two retrieval regimes. Our results show that retrieval quality is a dominant bottleneck for AI agents: when evidence is incomplete or skewed, all agents degrade substantially, and skill files cannot recover lost performance. Under high-quality retrieval, however, declarative skills consistently improve accuracy on procedural tasks and reduce orchestration errors, while the imperative state machine's brittleness does not reliably improve task success or compliance.
Abstract:We study orchestration mechanisms for tool-using AI agents in realistic customer-service workflows over an unstructured knowledge base. We argue that declarative agents -- AI agents equipped with natural-language skill files appended to the system prompt -- are an effective orchestration paradigm. Concretely, we compare (i) a DeclarativeAgent that reads three domain-specific skill files at inference time and decides its own control flow, (ii) an ImperativeAgent based on a programmatic state machine with explicit phases, and (iii) an unscaffolded baseline agent modeled after the $\tau$-Knowledge benchmark agent. Our ImperativeAgent is motivated by externalised-control inference as in Recursive Language Models and graph-based orchestration frameworks. We formalise the three agents as policy classes within a decentralised partially-observable Markov decision process and analyse their information-theoretic and structural properties; we then test the predicted differences empirically on five language models and two retrieval regimes. Our results show that retrieval quality is a dominant bottleneck for AI agents: when evidence is incomplete or skewed, all agents degrade substantially, and skill files cannot recover lost performance. Under high-quality retrieval, however, declarative skills consistently improve accuracy on procedural tasks and reduce orchestration errors, while the imperative state machine's brittleness does not reliably improve task success or compliance.
Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
arXiv:2606.06923 [cs.AI]
(or arXiv:2606.06923v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2606.06923 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 296 words
Related coverage
Manage agents, tools and skills at scale with AWS Agent Registry
AWS · Aug 31, 2026
New in Amazon Bedrock AgentCore: Build agents with broader knowledge and continuous learning
AWS · Jun 17, 2026
Performance Profiling on AMD GPUs – Part 5: Profiling-Driven Kernel Optimization with an AI Code-Assist Tool
AMD · Jul 16, 2026
A shared agentic platform for Wood Mackenzie, on Amazon Bedrock AgentCore
AWS · Sep 17, 2026