AI & ML
Software Engineering for LLM-Based Agents: A Lifecycle-Oriented Survey
arXiv cs.SE0 views
arXiv:2611.23260v1 Announce Type: new Abstract: Large language model (LLM)-based Agents perform tasks through tool use and interaction with external environments. Public engineering guidance commonly begins with a target task, selects a suitable workflow or autonomous Agent design, and refines prompts, tools, and execution strategies using task outcomes, latency, and cost. When delivered as software systems, Agents must also satisfy stakeholder requirements, action boundaries, quality checks, and maintenance needs. Model-directed execution can vary across runs, failures can span model decisions and tools, and changes to prompts, models, or interfaces can affect previously accepted behavior. This survey examines how LLM-based Agents are engineered across five software lifecycle stages: Requirements Engineering, Architecture and Design, Implementation and Integration, Testing and Assurance, and Maintenance and Evolution. Based on a main collection of 355 papers, we compare the engineering problems addressed, explain how methods and practices work, and examine their reported evaluations and limitations within each stage. We synthesize how existing research supports Agent development activities and identify open problems and research directions. The result is a lifecycle-oriented reference for understanding, selecting, and further developing methods for engineering Agent software.
Source:https://arxiv.org/abs/2611.23260PreviousControlled Natural Language for Requirements Specification: A Systematic Literature ReviewNextArtificial Intelligence in Finance and the Early-Career Skills Shift: A Structured Evidence Synthesis of Decision Delegation, Bank Deployments, and Business Education
Related posts
Im looking fr any free ai models / tools are there?, leaVE UR Lights💡
0dev.to
I checked 10 SEO agency websites for AI-search readiness. The real bugs were the boring ones.
0dev.to

How much GPU memory does an LLM need? The whole sum, worked out on real cards
0dev.to

Routing Upscaling Jobs by Output Size, Not Input Size
0dev.to
A ten-second Flutter pause with a visible Skip button
0dev.to UX

Thống kê lịch sử đối đầu Aston Villa - Sunderland: Đại diện Midlands áp đảo toàn diện
0dev.to Data