Is DevRag truly local and free?
Yes. It runs entirely on your machine, with no API calls or external services, so there are no usage costs.
D
DevRag - Lightweight Local RAG MCP Server
@tomohiro-owada
Basic information
About DevRag - Lightweight Local RAG MCP Server
Lightweight local RAG MCP server for semantic vector
DevRag - Lightweight Local RAG MCP Server config
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"devrag": {
"type": "stdio",
"command": "/usr/local/bin/devrag"
}
}
}DevRag - Lightweight Local RAG MCP Server tools
Tool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
We auto-extract tools from the README. The maintainer can list them under a ## Tools heading to populate this section.
Overview
What is DevRag - Lightweight Local RAG MCP Server?
DevRag is a lightweight, local-only MCP server that provides Retrieval-Augmented Generation (RAG) capabilities to MCP hosts (e.g., Claude Desktop). It performs semantic vector search over indexed markdown files, dramatically reducing token usage and search latency—all without any API dependencies or costs.
How to use DevRag - Lightweight Local RAG MCP Server?
DevRag is distributed as a single binary with no dependencies. To use it, launch the binary and configure your MCP host to point to it. Once running, the server exposes tools such as search, index_markdown, list_documents, delete_document, and reindex_document for managing and querying your local knowledge base.
Key features of DevRag - Lightweight Local RAG MCP Server
- 40x token reduction via semantic vector search
- 15x faster responses (~95ms search latency)
- Runs entirely locally – no API costs
- Multi-language support – Japanese and English
- Filtered search by directory and filename patterns
- Single binary with no external dependencies
Use cases of DevRag - Lightweight Local RAG MCP Server
- Augmenting an AI assistant with a local markdown knowledge base
- Indexing personal documentation or project wikis for fast retrieval
- Reducing token usage in RAG pipelines when using paid LLM APIs
- Multi-language RAG for teams working in Japanese and English
Frequently asked questions
What are the runtime requirements?
DevRag ships as a single binary with no dependencies—just download and execute.
Which languages does it support?
Indexing and search are available in both Japanese and English.
How do I add my documents?
Use the `index_markdown` tool to add a markdown file to the index, and `list_documents` to review indexed content.
What transport does it use?
The server follows the Model Context Protocol, which typically uses STDIO transport for host–server communication.
Are there any limits on document size or number?
The README does not specify explicit limits; performance and index capacity depend on the local environment.
More Memory & Knowledge MCP servers
LoreKeeper RPG (lorekeeper.com)
LoreKeeper LtdMess in, magic out: pour in your D&D and TTRPG notes and LoreKeeper keeps them as a private store for your world. Ask Claude or ChatGPT about it, or build a sourcebook.
Sensefold
SenseFold LabsPersonal context for every AI: search, read, and write back to your private Markdown library of articles, threads, PDFs, notes, and captured ChatGPT/Claude/Gemini/Grok conversations. Remote server, OAuth 2.1 paste-and-au
priostack
ideaswaveOpen-source Python SDK & MCP interface for Priostack Agent Context Network (ACN). Zero-setup, autonomous memory orchestration for AI agents.
PLUR
plur-aiAI agents start every session with amnesia — you re-explain the project, repeat your preferences, and correct the same mistakes over and over. PLUR gives them a memory that persists. Your agent's corrections, preference
226
Leviathan Systems Public Content MCP
Leviathan SystemsSearch published GPU deployment, rack integration and structured cabling resources; read canonical documents and Leviathan Systems company service scope. Anonymous, read-only Streamable HTTP MCP. No credentials required.
Comments