Newest papers

Selecting and Composing Agent Design Patterns: Industrial Case Evidence and Controlled Reference-System Experiments

Jia Huang, Bingsheng Ru, Bo Liang, Yuke Xiong, Haiyan Fan, Joey Tianyi Zhou · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.25629 · cited by 0 works (OpenAlex) · open-access PDF

German

Agentenrahmenkataloge beschreiben eher konventionelle Architekturalternativen, aber Ingenieure benötigen auch Beweise für ihre Auswahl und Komposition. Wir erweitern ein kognitives Funktions-Exekutive-Topologie-Rahmen mit einem evidenzgesteuerten Auswahlverfahren und ausdrücklichen Kompositionskontrakten. Die Studie kombiniert zwei industrielle Fallkonten, ein Beitrags-Kollaborationsdesign, sieben Praktiker-Workshops und ausführbare Fehler-Injektions-Experimente. Drei lokale Referenzsysteme verwenden SQLite-Status, gemeinsame Artefakte-Dateien und eine Datei-unterstützte Karte-Veröffentlichung-Surrogate. Eine 2 x 2-Ablation trennt funktionspezifische Verträge von Ausführungskontrollen über 36 Szenarienfamil

Machine translation (Meta M2M100); the English original follows.

Abstract

Agent pattern catalogs describe architectural alternatives, but engineers also need evidence for selecting and composing them. We extend a cognitive-function--execution-topology framework with an evidence-driven selection procedure and explicit composition contracts. The study combines two industrial case accounts, a contributor's collaboration design, seven practitioner workshops, and executable fault-injection experiments. Three local reference systems use SQLite state, shared artifact files, and a file-backed map-publication surrogate. A 2 x 2 ablation separates function-specific contracts from execution controls across 36 scenario families, with ten input variants per family. Under equal family weighting, composition correctly completes or safely refuses 11 of 12 families in each system, compared with 2, 2, and 1 for the simple baselines. These are constructed challenge-suite results, not production success rates. Recovery without consumer-side verification increases false-success cases in the map surrogate. All composed systems still fail a boundary case in which trusted evidence is wrong. An audit of a frozen six-arm language-model pilot reproduces its scores but confirms an evaluator-dependent ranking reversal. The contribution connects industrial failure, pattern selection, executable treatment, external acceptance, and revision. Selection is demonstrated rather than evaluated against an alternative procedure; catalog admission rests on separate criteria. The results support bounded composition mechanisms, not a claim that the framework improves engineers' design decisions over an equally informative conventional method.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Executable, Computer science, Selection (genetic algorithm), Audit, Artifact (error), Ranking (information retrieval)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange