From 897554b561ebdedd562991f7f600def51bc8ec26 Mon Sep 17 00:00:00 2001 From: fish827-08 <3264527905@qq.com> Date: Thu, 27 Aug 2026 21:51:10 +0800 Subject: [PATCH] Add rag-kb: local-first agent memory & knowledge service MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Adds **rag-kb** — a local-first, completely free agent memory & knowledge service. - Hybrid retrieval for Chinese & English: BGE-M3 dense vectors + BM25 (jieba), fused with RRF - Local-first & offline-capable: single process, embedded ChromaDB, no external services; retrieval works with the network cut off - Dual protocol: REST + native MCP server from the same process - Fully functional without any LLM; optional Ollama/cloud unlocks RAG answering - Apache-2.0, English README included Submitted by the project author. Happy to adjust the entry if needed. --- README.md | 1 + 1 file changed, 1 insertion(+) diff --git a/README.md b/README.md index 365340f64..a616d1434 100644 --- a/README.md +++ b/README.md @@ -2197,6 +2197,7 @@ Connect AI agents to industrial equipment, machinery, and operational technology Persistent memory storage using knowledge graph structures. Enables AI models to maintain and query structured information across sessions. +- [fish827-08/rag-kb](https://github.com/fish827-08/rag-kb) 🐍 🏠 🪟 - Local-first agent memory & knowledge service with hybrid retrieval (BGE-M3 dense vectors + BM25 via jieba, fused with RRF), document & webpage ingestion, and optional RAG answering via local Ollama. Single process, embedded ChromaDB, REST + native MCP server; fully functional offline without any LLM. Apache-2.0. - [AkashGoenka/coldstart](https://github.com/AkashGoenka/coldstart) [![AkashGoenka/coldstart MCP server](https://glama.ai/mcp/servers/AkashGoenka/coldstart/badges/score.svg)](https://glama.ai/mcp/servers/AkashGoenka/coldstart) 📇 🏠 🍎 🪟 🐧 - Codebase memory for coding agents, with no embeddings and no API key. A deterministic AST index answers "which files are relevant to this task?" in milliseconds, and agents write durable notes about the repo that are content-hash checked — a note flags itself stale the moment the code it describes changes. Notes are markdown inside the repo, so they commit and review alongside your code. `npm i -g @cstart/coldstart && coldstart init` - [jayasukuv11-beep/agenthelm](https://github.com/jayasukuv11-beep/agenthelm) [![jayasukuv11-beep/agenthelm MCP server](https://glama.ai/mcp/servers/jayasukuv11-beep/agenthelm/badges/score.svg)](https://glama.ai/mcp/servers/jayasukuv11-beep/agenthelm) 📇 🐍 ☁️ 🏠 🍎 🪟 🐧 - Shared, versioned memory and governance control plane for AI coding agents. Compiler pipeline resolves architectural decision conflicts across Claude Code, Cursor, and custom agent fleets. `npx -y agenthelm-mcp` - [jmpdsevilla/BovedIA](https://github.com/jmpdsevilla/BovedIA) [![jmpdsevilla/BovedIA MCP server](https://glama.ai/mcp/servers/@jmpdsevilla/BovedIA/badges/score.svg)](https://glama.ai/mcp/servers/@jmpdsevilla/BovedIA) 📇 🏠 🍎 🪟 🐧 - Personal memory for Claude Code in plain Markdown notes you own. Doesn't load context blindly: a router note decides what to load and when, plus an "alma" (soul) layer for what you think and feel, not just tasks. `npx bovedia`