// the find
Andrew-Jang/RAGHub
A community-driven collection of RAG (Retrieval-Augmented Generation) frameworks, projects, and resources. Contribute and explore the evolving RAG ecosystem.
RAGHub is a link directory for the RAG ecosystem — frameworks, evaluation tools, engines, data prep libraries, and other resources organized as markdown tables. There's no code here; it's a reference page for people trying to figure out which RAG tool to reach for, run as a community list for r/RAG.
The FAQ section actually earns its place: framework-vs-engine distinction, a vector DB comparison table, and a table of common RAG failure modes with fixes are genuinely useful orientation for someone new to the space. It's broad — frameworks, evaluators, engines, and data-prep tools are split into separate sections instead of one undifferentiated pile. Contribution flow is a plain PR against a markdown table, which is about as low-friction as it gets.
It's a list, so the maintenance burden is unbounded and there's no visible process for pruning dead or renamed projects — entries point to things like an org with no repo link, or a website with '-' for GitHub, which suggests link rot is already creeping in. The 'Activity' column is largely meaningless (values like '1h ago' or '3h ago' across dozens of entries in the sample look auto-generated at scrape time rather than reflecting real commit activity, so it can't be trusted for judging whether a project is alive). No filtering by license, language, or actual maturity beyond a stars badge, so you still have to click through everything yourself. With 2000+ stars driving a lot of inbound PRs, there's no evident curation bar — anything with a landing page and a description seems to get a row, including entries that look like self-promotion (vague one-line pitches, no GitHub link at all).