A curated weekly list of AI agent research papers from arXiv for AI/LLM agent developers.
Add your paper to the relevant category in README.md:
| **[Paper Title](https://arxiv.org/pdf/XXXX.XXXXvX)** - 1–2 sentence description. |Add to the matching category:
- Tooling — frameworks, training, optimization, planning, context engineering, architecture
- Memory & RAG — memory systems, knowledge retrieval, RAG pipelines, procedural memory
- Multi-Agent — coordination, communication, orchestration, collaboration, negotiation
- Eval & Observability — benchmarks, testing, monitoring, debugging, tracing, explainability
- Security — attacks, defenses, alignment, trust, prompt injection, safety
- Paper must be published on arXiv
- Paper must be relevant to AI/LLM agent developers
- Description must be 1–2 sentences, plain English, no academic phrasing
- Assign exactly one category per paper
- Link must point to the PDF URL
- Write like you're explaining to a developer friend — no academic language
- Explain what the research is about, not why it's important
- No "helping developers" product-pitch tone
- No editorial commentary or personal opinions
- Numbers need context (e.g., "93% precision on function-level TTP recognition")
Add paper: Paper Title
Weekly lists go under lists/ with the naming convention awesome-agents-YYYY-MM-DD-to-YYYY-MM-DD.md.
- Verify your links work before submitting
- We review all submissions and may decline papers that don't fit the AI agent ecosystem
- Check existing issues and PRs first
- Open a new issue for questions