awesome-ai-agent-papers

Why we included this project

Keeping up with AI agent research means sorting through hundreds of arXiv preprints every week, and most of them are only loosely related to the field. This list does that sorting for you: it tracks papers published from January 2026 onward, filters for ones that genuinely touch the agent ecosystem, and groups them by theme such as multi-agent coordination, memory and RAG, tooling, evaluation, and security. Each entry includes a short plain-language summary of what the paper proposes and why it matters, so you can skim a section and decide which full papers are worth reading. The list is refreshed weekly, which matters in a fast-moving field. For an engineer deciding what to adopt or a researcher scouting new architectures, it's a practical starting point rather than something you deploy.