loci runs on your own machine — the client starts it, so there is no endpoint to ping. 480 installs a week from pypi. Last commit 1 Nov 2026.
Cross-session memory and hybrid RAG over your notes and docs. Local-first, with cited answers.
Today is the operative word: we check loci every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add loci -- uvx loci-rag
{
"mcpServers": {
"loci": {
"command": "uvx",
"args": [
"loci-rag"
]
}
}
}
[mcp_servers.loci]
command = "uvx"
args = ["loci-rag"]
{
"mcpServers": {
"loci": {
"command": "uvx",
"args": [
"loci-rag"
]
}
}
}
{
"servers": {
"loci": {
"type": "stdio",
"command": "uvx",
"args": [
"loci-rag"
]
}
}
}
Local-first RAG MCP server: hybrid search over a folder of your own documents
Permission-aware RAG for agents — ingest docs, hybrid search, cited answers (hosted API).
Local-first AI memory layer with hybrid search. Postgres + pgvector. Self-hosted, MIT.
Build a RAG pipeline station by station and search your own documents with cited answers.
Lightweight self-hosted RAG knowledge base with hybrid retrieval and cited answers.
Local-first agentic RAG with citations - hybrid search, reranking, multimodal document retrieval
Local RAG MCP server with hybrid search, PDF/DOCX support, and zero-config setup
Shared memory + orchestration for your coding agents. Local-first MCP, vector RAG.
Answers built from our own checks of this server.