📡 MCP Radar

do-md/homekb

Local-first personal Markdown knowledge base with semantic search, AI Q&A, MCP, desktop/web clients, and a self-hostable zero-data-at-rest relay.

6 stars
0 forks
TypeScript
momentum ▲ 12.0
created 2026-07-17
on radar since 2026-07-23
aichatgptclaudecloudflare-workersknowledge-managementlocal-firstmarkdownmcpmodel-context-protocolnextjspersonal-knowledge-baseprivacyragrustsecond-brainself-hostedsemantic-searchsqlitesqlite-vectauri
View on GitHub ↗Homepage ↗

About homekb

A personal Markdown knowledge base that lives on your computer and stays within reach of your AI — locally or from anywhere.

- Files first. Your notes remain ordinary .md files in a folder you control. - Agent native. Claude Code, Codex, Claude, ChatGPT, and other MCP clients can search, read, create, update, and share notes. - Local by construction. The index, retrieval, and writes stay on your home computer; AI calls use the provider you configure. - Remote without an account. Pair with a one-time code; the relay stores relationships and token hashes, never your knowledge-base content at rest. - Two commands to your phone. Install the engine, run homekb pair, and finish setup in the Web UI.

HomeKB turns a folder of Markdown files into a semantic knowledge base without moving ownership into a cloud app.

Drop notes into ~/.homekb/notes/ — or point HomeKB at an existing Markdown folder — and the Rust engine incrementally builds a local sqlite-vec index. You can then search by meaning, ask questions with citations, edit notes, or let an AI agent work with the library through MCP.

The command-line engine is the product core. The desktop app and Web UI are renderers over the same RPC contract, while the relay is only a pipe between a remote client and the engine running at home.

- Compile Markdown into summaries, chunks, document types, suggested questions, and embeddings. - Retrieve with dual-pool KNN over document summaries and chunks, fused with RRF. - Switch to whole-category enumeration when a question asks for coverage rather than a top-K match. - Answer questions fr

From the project README.

Maintaining this server?

Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:

[![On MCP Radar](https://mcp.liqiwa.com/badge.svg)](https://mcp.liqiwa.com/s/do-md--homekb.html)

Related MCP servers

bybit-exchange/kaas

Turn scattered notes, docs and transcripts into a queryable Markdown wiki — an LLM knowledge-base compiler with MCP access, no embeddings, self-hosted.

⭐ 82Python
docouno/notarium

File-first knowledge base for humans and AI agents — self-hosted Markdown notes with a web editor and a built-in MCP server.

⭐ 3TypeScript
menot-you/n-memory

Memory your agent can trust — every answer carries its source; when it doesn't know, it says so. Hermetic, local, MCP.

⭐ 5Rust
xfloukiex-lab/pandaclip

One local-first MCP server for agent working state — clipboard history & snippets, TTL cache, file-organizer overlay, and a knowledge graph — plus a live desktop activity lens. 40 tools, SQLite everyw

⭐ 4TypeScript
rwnalds/engram

Self-hosted MCP server + dashboard giving Claude Code, Cursor & Hermes agents shared long-term memory over a git-backed markdown vault. Agent-native, Obsidian-compatible second brain — markdown RAG, n

⭐ 4TypeScript
MuhittinYilmazer/akana

Self-hosted, local-first AI assistant server: swappable LLM providers, review-gated memory, an encrypted vault, and wake-word voice. No accounts, no telemetry.

⭐ 3Python

📬 Get the weekly radar in your inbox

The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.