Awesome AI Agent Papers
A hand-picked, weekly-updated collection of AI agent research papers published in 2026, sourced from arXiv and organized by topics such as multi-agent coordination, memory and RAG, tooling, evaluation and security.
Awesome AI Agent Papers is a curated collection of AI agent research papers published in 2026 and sourced from arXiv, covering multi-agent coordination, memory and retrieval-augmented generation, agent tooling, evaluation and observability, and AI agent security. Only papers published from January 2026 onward are included, and the list is updated weekly as new papers are filtered and categorized.
Entries summarize each paper's contribution, for example an autonomous, cross-disciplinary research system built around minimizing prompt-engineering overhead, or a persona-conditioned method for using LLM agents as synthetic users in UI and UX evaluation. Papers are grouped into named sections such as multi-agent, memory and RAG, evaluation and observability, agent tooling, and AI agent security, each with a running paper count.
It is maintained on GitHub under the MIT license and accepts community contributions.