multi-agent · research · arXiv
Agentic arXiv creator
Multi-agent system that discovers, evaluates, and ranks arXiv papers, then generates LinkedIn posts from top findings.
Overview
An autonomous research system that uses a team of specialized AI agents to search arXiv for cutting-edge papers, evaluate them against domain-specific rubrics, and optionally generate polished LinkedIn posts from the top-ranked findings. Built on Anthropic Claude and LangGraph.
What it does
- An LLM agent translates natural-language queries into structured execution plans (keywords, date ranges, search strategies).
- Searches arXiv against domain-specific rubrics.
- Evaluates and ranks paper candidates.
- Generates publish-ready LinkedIn posts from the top-ranked findings.
Stack
Anthropic Claude · LangGraph · arXiv API · Gradio.
Status
Open source (MIT). Live on Hugging Face Spaces.