CoexistAI
Automates and simplifies diverse research workflows by integrating large language models with web search, social media, mapping, and code exploration.
概要
CoexistAI is a modular, developer-friendly framework designed to streamline and automate complex research workflows. It empowers users to build, search, summarize, and automate research tasks by integrating with a wide array of data sources, including web search (via SearxNG), Reddit, YouTube, mapping tools, and GitHub or local code repositories. The framework supports pluggable LLMs and embedders, ensuring high performance through asynchronous and parallel execution. It can be used as a Python library, accessed via a FASTAPI, or integrated seamlessly with MCP servers, making it highly versatile for various research and prototyping needs.
主な機能
- GitHub & Local Codebase Analysis to explore, summarize, explain, and answer questions about repositories.
- YouTube Search & Summarization, allowing search by query or URL with custom prompts.
- Map & Location/Route Search to find places, generate routes, and identify points of interest.
- 83 GitHub stars
- Comprehensive Web Explorer for querying, summarizing, and extracting context from web content.
- Advanced Reddit Deep Dive for custom searches, sorting, filtering, and top comment retrieval.
ユースケース
- Summarizing articles, research papers, YouTube videos, and Reddit discussions with AI-driven insights.
- Automating comprehensive research across multiple online sources for academic or personal projects.
- Analyzing and understanding codebases from GitHub or local repositories to quickly grasp project structure and functionality.