Safety
Humans are Missing from AI Coding Agent Research
arXiv:2608.12355v1 Announce Type: cross Abstract: Recent progress in AI coding agent research has led to rapid improvements in agents' ability to autonomously perform complex software engineering task
arXiv:2608.12355v1 Announce Type: cross Abstract: Recent progress in AI coding agent research has led to rapid improvements in agents' ability to autonomously perform complex software engineering tasks, from editing large codebases to executing long-horizon development workflows. As these systems make strides, however, the primary bottleneck to practical usefulness increasingly shifts away from pure task-solving capability, and toward challenges in how users communicate with, supervise, and trust agents. In this position paper, we argue for a reorientation from autonomous to human-centered coding agents: systems designed not only to complete tasks, but to collaborate effectively with people. We identify four core interaction-level dimensions that characterize the human-agent task-solving loop: task alignment, verifiability, steerability, and adaptability. Finally, we outline concrete research directions to advance these dimensions, including user-involved coding environments, comprehensive verification mechanisms, and principled measures of human-agent interaction quality.
Related
- Socratic-SWE: Self-Evolving Coding Agents via Trace-Derived Agent Skills
- CODESKILL: Learning Self-Evolving Skills for Coding Agents
- Debugging the Debuggers: Failure-Anchored Structured Recovery for Software Engineering Agents
- Execution-Grounded Security Testing for Coding Agents in Software Engineering Pipelines
- Lingering Authority: Revocable Resource-and-Effect Capabilities for Coding Agents
Source: arXiv cs.AI | 2026-08-14