A large, active GitHub list of must-read papers on LLM agents, with extensive indexing of papers across agent categories and memory/planning topics. Includes numerous paper entries and a structured content outline.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
LLMAgentPapers is a curated list of papers on large language model (LLM) agents, organized into sections such as Overview, Agent (with subtopics Personality, Memory, Planning, Tool use, RL training), Multiple Agents (Task-Oriented Communication, Collaborative Exchanges, Adversarial Interactions, Casual/Open Conversations), Application, Framework, and Other resources. The repository provides a content-focused compilation rather than code or runtime tooling.
How it works
The repository aggregates papers and curates them into themed subsections, presenting each entry with title, authors, and arXiv/abs links. It uses a structured outline in the README to categorize papers by topic (e.g., Personality, Memory, Planning) and by agent interactions (single vs multiple agents) and applications. The README shows a long list under each subsection without executable code.
Getting started
No installation or runtime setup is required beyond viewing the README. There are no explicit installation instructions in the truncated README text provided. The repository description indicates a focus on listing papers rather than providing tooling.
Recent releases
The RELEASES section indicates 'latest 0' with 'none' listed, meaning there are no tagged releases in this repository.
Traction
stars_7d and stars_1d are not provided in the FACTS block, but overall stars are listed as 3097 and forks as 185. No daily or 7-day star activity is shown in the provided data, so the Traction section is omitted.
Behind the repo
There is no linked startup or company information provided in the FACTS. The repository appears to be an academic/curated list rather than a company-backed project.
Caveats
License is listed as MIT in the README badge, but the FACTS indicate license: none listed. The README snippet shows a LICENSE badge stating MIT, but the factual license field in the SUMMARY data is not explicit. There is no explicit age aside from created: 2023-05-19 and last_push: 2026-07-27. The content is a lengthy bibliography rather than code, with no runtime dependencies.