rmcp_memex
by LibraxisAI
Overview
Provides a local Retrieval-Augmented Generation (RAG) toolset backed by an embedded LanceDB vector store and local embeddings, serving via JSON-RPC.
Installation
cargo run --release -- --log-level infoEnvironment Variables
- DISABLE_MLX
- DRAGON_BASE_URL
- MLX_JIT_MODE
- MLX_JIT_PORT
- EMBEDDER_PORT
- RERANKER_PORT
- EMBEDDER_MODEL
- RERANKER_MODEL
- FASTEMBED_CACHE_PATH
- HF_HUB_CACHE
- LANCEDB_PATH
- PROTOC
Security Notes
The `rag_index` function takes a file path as an RPC argument. Without explicit sanitization or restriction to a designated directory, this could be vulnerable to path traversal, allowing a malicious client to index or read arbitrary files on the host system (e.g., `/etc/passwd`). The MLX bridge makes configurable HTTP calls, but defaults to localhost. No obvious 'eval' or hardcoded secrets were found.
Similar Servers
haiku.rag
An opinionated agentic RAG system that uses LanceDB for vector storage, Pydantic AI for multi-agent workflows, and Docling for document processing, exposing its capabilities as MCP tools for AI assistants.
mcp-local-rag
Provides a local RAG-like web search capability for LLMs through the Model Context Protocol without external APIs.
Context-Engine
A Retrieval-Augmented Generation (RAG) stack for codebases, enabling context-aware AI agents for developers and IDEs through unified code indexing, hybrid search, and local LLM integration.
rag-server-mcp
Provides Retrieval Augmented Generation (RAG) capabilities to Model Context Protocol (MCP) clients by indexing local project documents and retrieving relevant information for LLMs.