Multi-agent AI (LangGraph) to audit AI papers for reproducibility

PaperPulse, a browser-integrated multi-agent AI system built on a LangGraph orchestrator, audits scientific papers, news articles and web content for methodological rigor, statistical validity and code-to-paper alignment. The system runs four specialized agents in sequence — a Universal Extractor, a GitHub/Code Grounder, a Statistical Critic and a Citation-Grounded Synthesizer — behind a FastAPI backend, with a Chrome extension sending the active tab's URL and rendering the resulting structured JSON report as Markdown. PaperPulse positions itself against generic "Chat with PDF" wrappers by cross-referencing claims, searching GitHub for official repositories and flagging reproducibility gaps with exact citations.

PaperPulse is a browser-integrated, multi-agent AI system that automatically audits scientific papers, news articles, and web content for methodological rigor, statistical validity, and code-to-paper alignment. Unlike generic "Chat with PDF" wrappers, PaperPulse orchestrates specialized AI agents to cross-reference claims, search GitHub for official repositories, and flag reproducibility gaps with exact citations. php graph TD A Chrome Extension: Active Tab -- |Universal URL| B FastAPI Backend B -- C{LangGraph Orchestrator} C -- D Agent 1: Universal Extractor D -- E Agent 2: GitHub/Code Grounder E -- F Agent 3: Statistical Critic F -- G Agent 4: Citation-Grounded Synthesizer G -- H Structured JSON Report H -- |Rendered Markdown| A style C fill: 4f46e5,stroke: 312e81,stroke-width:2px,color: fff style A fill: 0f172a,stroke: 334155,stroke-width:2px,color: fff