RAG & Vector Databases: Fastest-Growing Projects — May 05, 2026
Today's RAG & Vector Databases, we're seeing a surge of interest in tools that enable local-first knowledge bases and hybrid search capabilities. Developers are increasingly looking for ways to turn raw research into organized, interlinked knowledge graphs that can compound over time. This trend is reflected in the growth scores of several repositories on our radar.
Swarmclawai's swarmvault has taken the top spot with a growth score of 18.83 and 375 stars. This local-first RAG knowledge base compiler allows users to turn raw research into a persistent markdown wiki, knowledge graph, and hybrid search that compound over time. Its high growth score can be attributed to its unique approach to compiling knowledge bases and its support for popular AI models like Claude Code and Codex.
Rolandpg's zettelforge is another repository gaining traction, with a growth score of 11.22 and 33 stars. This Python-based tool provides agentic memory for CTI, featuring STIX knowledge graphs, threat-actor alias resolution, offline-first RAG, and MCP server support for Claude Code and LangChain agents. Its growth can be attributed to its focus on security analysts and the need for intelligent threat intelligence analysis tools.
Nashsu's llm_wiki has maintained a strong presence with a growth score of 6.85 and an impressive 5,750 stars. This cross-platform desktop application turns documents into an organized, interlinked knowledge base automatically, using LLMs to incrementally build and maintain a persistent wiki from sources. Its enduring popularity can be attributed to its user-friendly approach to creating knowledge bases and its ability to work with various document formats.
Ais1on's CTI-RAG has a growth score of 5.71 and 223 stars, despite having no commits in the last 30 days. This Retrieval-Augmented Generation framework for Cyber Threat Intelligence integrates knowledge graph and causal reasoning capabilities to provide security analysts with an intelligent threat intelligence analysis tool. Its slow growth may be due to its niche focus on CTI, but it remains a valuable resource for security professionals.
Lastly, yanhua1010's zero-to-ai-fullstack has a growth score of 3.83 and 150 stars. This repository documents a Java backend engineer's journey learning AI full-stack in public, covering topics like Python, FastAPI, RAG, pgvector, and Next.js. Its growth can be attributed to the increasing demand for AI knowledge among developers and the value of learning from someone else's experiences.
Overall, Today's trends in RAG & Vector Databases highlight the growing interest in local-first knowledge bases, hybrid search capabilities, and intelligent threat intelligence analysis tools. As these repositories continue to grow, we can expect to see more innovative solutions emerging in the space.
Swarmclawai's swarmvault has taken the top spot with a growth score of 18.83 and 375 stars. This local-first RAG knowledge base compiler allows users to turn raw research into a persistent markdown wiki, knowledge graph, and hybrid search that compound over time. Its high growth score can be attributed to its unique approach to compiling knowledge bases and its support for popular AI models like Claude Code and Codex.
Rolandpg's zettelforge is another repository gaining traction, with a growth score of 11.22 and 33 stars. This Python-based tool provides agentic memory for CTI, featuring STIX knowledge graphs, threat-actor alias resolution, offline-first RAG, and MCP server support for Claude Code and LangChain agents. Its growth can be attributed to its focus on security analysts and the need for intelligent threat intelligence analysis tools.
Nashsu's llm_wiki has maintained a strong presence with a growth score of 6.85 and an impressive 5,750 stars. This cross-platform desktop application turns documents into an organized, interlinked knowledge base automatically, using LLMs to incrementally build and maintain a persistent wiki from sources. Its enduring popularity can be attributed to its user-friendly approach to creating knowledge bases and its ability to work with various document formats.
Ais1on's CTI-RAG has a growth score of 5.71 and 223 stars, despite having no commits in the last 30 days. This Retrieval-Augmented Generation framework for Cyber Threat Intelligence integrates knowledge graph and causal reasoning capabilities to provide security analysts with an intelligent threat intelligence analysis tool. Its slow growth may be due to its niche focus on CTI, but it remains a valuable resource for security professionals.
Lastly, yanhua1010's zero-to-ai-fullstack has a growth score of 3.83 and 150 stars. This repository documents a Java backend engineer's journey learning AI full-stack in public, covering topics like Python, FastAPI, RAG, pgvector, and Next.js. Its growth can be attributed to the increasing demand for AI knowledge among developers and the value of learning from someone else's experiences.
Overall, Today's trends in RAG & Vector Databases highlight the growing interest in local-first knowledge bases, hybrid search capabilities, and intelligent threat intelligence analysis tools. As these repositories continue to grow, we can expect to see more innovative solutions emerging in the space.