source-citations

长篇深度研究报告的调研/写作助手，一句话自动跑完完整报告输出高质量 PDF 报告

A sophisticated RAG (Retrieval-Augmented Generation) Telegram bot that transforms articles and documents into interactive knowledge bases. Upload PDFs/URLs and get AI-powered answers with source citations.

Search infrastructure for AI agents: free by default, MCP-ready, LLM-context aware, and production-grade when needed.

Local-first MCP memory server for AI agents, with source citations and OKF/Markdown knowledge bases.

Rust MCP server for source-cited search over a local Zoltan Pozsar macro/liquidity corpus.

MCP search and evidence tool for AI agents with query rewriting, source zoom-in, sourced answers, and runtime metrics.

A simple RAG toolkit.

Memtruth SDK: evidence, parse, corpus contracts, chunking, projection, and diagnostics for AI applications.

A LangGraph-based multi-agent PDF analysis system with RAG, dynamic planning, collaboration, lightweight reflection/retry, memory, source citations, and a Streamlit UI.

AI-assisted PDF/DOCX packet structuring workflow with source citations, semantic retrieval, deterministic validation, and reviewer-facing run sheets.

RAG-powered document assistant — upload PDFs, chat with them using BM25 retrieval + LLM, and get answers with verified source citations. Built with Next.js 16, Prisma, and a custom retrieval pipeline.

Cite Nook Ai is a minimal local-first RAG application built with React, FastAPI, Ollama, PostgreSQL/pgvector, and Docker. Upload persistent documents, track indexing status, choose chat and embedding models, and ask grounded questions with source citations.

Provider-neutral attribution, source-footer, and royalty settlement system for AI outputs.

Turn a local problem into a source-cited advocacy website with a guided research and editorial workflow.

Repository for DevVault AI web application.