Retrieval-Conditioned Topology Selection with Provable Budget Conservation for Multi-Agent Code Generation

TL;DR

This paper introduces RGAO, a retrieval-guided architecture for multi-agent code generation that adaptively selects orchestration topologies based on code complexity, ensuring budget conservation with provable guarantees.

Contribution

It combines complexity-conditioned routing and formal resource algebras to enable provable budget conservation in dynamic topology selection for multi-agent code generation.

Findings

Reduced misrouting from 30.1% to 8.2% using the topology router.

Achieved sub-millisecond DAG construction and linear scalability.

Demonstrated empirical effectiveness of the hierarchical code retrieval engine.

Abstract

Multi-agent LLM systems for code generation face a fundamental routing problem: the optimal orchestration topology depends on the structural complexity of the code under modification, yet existing systems select topologies without consulting the codebase. We present Retrieval-Guided Adaptive Orchestration (RGAO), an architecture that closes this loop by extracting a structural complexity vector from a hierarchical code index before selecting the orchestration topology. RGAO operates within Code-Agent, a multi-agent framework whose sub-agents are governed by formal contracts with six-dimensional budget vectors. Our headline contribution is the composition of two previously separate lines of work -- complexity-conditioned LLM routing and formal resource algebras -- yielding a property neither admits alone: provable budget conservation under retrieval-conditioned dynamic topology selection. Concretely we contribute: (1) a complexity-conditioned topology router that reduces proxy-measured misrouting from 30.1% to 8.2%; (2) a budget algebra with a structural-induction conservation theorem; and (3) a hierarchical code retrieval engine. Empirical evaluation demonstrates sub-millisecond DAG construction and linear tree-index scalability.