Inside the Scaffold: A Source-Code Taxonomy of Coding Agent Architectures

• URL: http://arxiv.org/abs/2604.03515v1
• Date: Fri, 03 Apr 2026 23:30:02 GMT
• Title: Inside the Scaffold: A Source-Code Taxonomy of Coding Agent Architectures
• Abstract summary: LLM-based coding agents can localize bugs, generate patches, and run tests with diminishing human oversight.<n>The scaffolding code that surrounds the language model (the control loop, tool definitions, state management, and context strategy) remains poorly understood.<n>This paper presents a source-code-level architectural taxonomy derived from analysis of 13 open-source coding agent scaffolds.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: LLM-based coding agents can localize bugs, generate patches, and run tests with diminishing human oversight, yet the scaffolding code that surrounds the language model (the control loop, tool definitions, state management, and context strategy) remains poorly understood. Existing surveys classify agents by abstract capabilities (tool use, planning, reflection) that cannot distinguish between architecturally distinct systems, and trajectory studies observe what agents do without examining the scaffold code that determines why. This paper presents a source-code-level architectural taxonomy derived from analysis of 13 open-source coding agent scaffolds at pinned commit hashes. Each agent is characterized across 12 dimensions organized into three layers: control architecture, tool and environment interface, and resource management. The analysis reveals that scaffold architectures resist discrete classification: control strategies range from fixed pipelines to Monte Carlo Tree Search, tool counts range from 0 to 37, and context compaction spans seven distinct strategies. Five loop primitives (ReAct, generate-test-repair, plan-execute, multi-attempt retry, tree search) function as composable building blocks that agents layer in different combinations; 11 of 13 agents compose multiple primitives rather than relying on a single control structure. Dimensions converge where external constraints dominate (tool capability categories, edit formats, execution isolation) and diverge where open design questions remain (context compaction, state management, multi-model routing). All taxonomic claims are grounded in file paths and line numbers, providing a reusable reference for researchers studying agent behavior and practitioners designing new scaffolds.

Related papers

• DocOps: A Verifiable Benchmark for Autonomous Agents in Complex Document Operations [106.74348504930124]
  We introduce DocOps, a deterministically verifiable evaluation framework underpinned by a hierarchical taxonomy.<n>Based on DocOps, we systematically evaluate representative closed- and open-source models across various agentic harnesses.<n>Our work exposes the capability boundaries of agents in maintaining global document consistency, shedding light on the future design of robust, non-destructive agents.
  arXiv  Detail & Related papers  (2026-07-22T07:52:59Z)
• Agentic System as Compressor: Quantifying System Intelligence in Bits [3.6471867330194008]
  Large language models are turning from isolated predictors into agentic systems.<n>We adopt an analytical viewpoint based on "compression is intelligence"<n>We operationalize the measure with arithmetic coding, seed coding, and a fallback, and evaluate it in five settings.
  arXiv  Detail & Related papers  (2026-06-24T15:32:49Z)
• How to Interpret Agent Behavior [56.59836196946289]
  We introduce ACT*ONOMY, a taxonomy for describing and analyzing agent behavior at runtime.<n>By providing a shared vocabulary, ACT*ONOMY helps researchers, agent designers, and end users interpret agent behavior more consistently.
  arXiv  Detail & Related papers  (2026-05-13T14:52:40Z)
• OrgAgent: Organize Your Multi-Agent System like a Company [75.47076168155817]
  We introduce OrgAgent, a company-style hierarchical multi-agent framework.<n>We find that multi-agent systems organized in a company-style hierarchy generally outperform other organizational structures.
  arXiv  Detail & Related papers  (2026-04-01T15:21:14Z)
• A Reference Architecture for Agentic Hybrid Retrieval in Dataset Search [0.6299766708197881]
  Ad hoc dataset search requires matching underspecified natural-language queries against sparse, heterogeneous metadata records.<n>We propose a bounded, auditable reference architecture for agentic hybrid retrieval that combines BM25 lexical search with dense-embedding retrieval.<n>An offline metadata augmentation step is introduced to reduce the vocabulary mismatch between user intent and provider-authored metadata.
  arXiv  Detail & Related papers  (2026-03-28T22:56:57Z)
• Improving Coherence and Persistence in Agentic AI for System Optimization [9.443037059325086]
  Engram organizes exploration into a sequence of agents that iteratively design, test, and analyze mechanisms.<n>We find that Engram exhibits superior performance across diverse domains including multi-cloud multicast inference, LLM request routing, and optimizing KV cache reuse in databases with natural language queries.
  arXiv  Detail & Related papers  (2026-03-22T17:04:50Z)
• A Two-Dimensional Framework for AI Agent Design Patterns: Cognitive Function and Execution Topology [54.72352855252792]
  We propose a two-dimensional classification that combines a Cognitive Function axis with seven categories and an Execution Topology axis with six structural archetypes.<n>The resulting 7x6 matrix identifies 27 named patterns, 13 with original names and provides a principled, framework-neutral, and model-agnostic vocabulary for AI agent architecture design.
  arXiv  Detail & Related papers  (2026-03-16T04:01:01Z)
• Architecture-Aware Multi-Design Generation for Repository-Level Feature Addition [53.50448142467294]
  RAIM is a multi-design and architecture-aware framework for repository-level feature addition.<n>It shifts away from linear patching by generating multiple diverse implementation designs.<n>Experiments on the NoCode-bench Verified dataset demonstrate that RAIM establishes a new state-of-the-art performance.
  arXiv  Detail & Related papers  (2026-03-02T12:50:40Z)
• Theory of Code Space: Do Code Agents Understand Software Architecture? [0.0]
  Code agents excel at isolated tasks yet struggle with multi-file software engineering requiring architectural understanding.<n>We introduce Theory of Code Space (ToCS), a benchmark that evaluates whether agents can construct, maintain, and update coherent architectural beliefs during AI exploration.
  arXiv  Detail & Related papers  (2026-02-28T11:40:17Z)
• A Hierarchical Multi-Agent System for Autonomous Discovery in Geoscientific Data Archives [0.0]
  PANGAEA-GPT is a hierarchical multi-agent framework designed for autonomous data discovery and analysis.<n>Unlike standard Large Language Model (LLM) wrappers, our architecture implements a centralized Supervisor-Worker topology.<n>We demonstrate the system's capacity to execute complex, multi-step deterministic runtime with minimal human intervention.
  arXiv  Detail & Related papers  (2026-02-24T20:37:38Z)
• LogicLens: Leveraging Semantic Code Graph to explore Multi Repository large systems [0.2519906683279152]
  We introduce LogicLens, a reactive conversational agent that assists developers in exploring complex software systems.<n>We present the architecture of the system, discuss emergent behaviors, and evaluate its effectiveness on real-world multi-repository scenarios.
  arXiv  Detail & Related papers  (2026-01-15T15:35:23Z)
• OctoBench: Benchmarking Scaffold-Aware Instruction Following in Repository-Grounded Agentic Coding [57.39403818250357]
  We introduce OctoBench, which benchmarks scaffold-aware instruction following in repository-grounded agentic coding.<n> OctoBench includes 34 environments and 217 tasks instantiated under three scaffold types, and is paired with 7,098 objective checklist items.<n>Experiments reveal a systematic gap between task-solving and scaffold-aware compliance, underscoring the need for training and evaluation.
  arXiv  Detail & Related papers  (2026-01-15T12:36:08Z)
• Designing Domain-Specific Agents via Hierarchical Task Abstraction Mechanism [61.01709143437043]
  We introduce a novel agent design framework centered on a Hierarchical Task Abstraction Mechanism (HTAM)<n>Specifically, HTAM moves beyond emulating social roles, instead structuring multi-agent systems into a logical hierarchy that mirrors the intrinsic task-dependency graph of a given domain.<n>We instantiate this framework as EarthAgent, a multi-agent system tailored for complex geospatial analysis.
  arXiv  Detail & Related papers  (2025-11-21T12:25:47Z)
• AgentSwift: Efficient LLM Agent Design via Value-guided Hierarchical Search [58.98450205734779]
  Large language model (LLM) agents have demonstrated strong capabilities across diverse domains.<n>Existing agent search methods suffer from three major limitations.<n>We introduce a comprehensive framework to address these challenges.
  arXiv  Detail & Related papers  (2025-06-06T12:07:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.