# ChrisGVE/workspace-qdrant-mcp [Health: Active]

**Category:** 🧠 Knowledge & Memory  
**Repository:** https://github.com/ChrisGVE/workspace-qdrant-mcp  
**GitHub Stars:** 4  
**Views:** 2  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/chrisgve-workspace-qdrant-mcp

## Description
Project-scoped semantic workspace memory for AI coding assistants. Watches your project files, auto-indexes code and docs into Qdrant with tree-sitter semantic chunking, LSP integration, and hybrid search (dense + sparse + RRF). 6 MCP tools: store, search, retrieve, grep, list, rules. Alpha — testers and feedback welcome.

## Tools
Capabilities this server exposes over MCP:

- **search** — Hybrid semantic + keyword search across indexed content
- **retrieve** — Direct document lookup by ID or metadata filter
- **rules** — Manage persistent behavioral rules
- **store** — Store content, register projects, save notes
- **grep** — Exact substring or regex search using FTS5
- **list** — List project files and folder structure

## Claude Desktop Quick Installation
No confirmed install details yet, so no config is published. Follow the setup steps in the repository above.

## Documentation

## What ChrisGVE/workspace-qdrant-mcp MCP server does

ChrisGVE/workspace-qdrant-mcp MCP server provides project-scoped memory for AI coding assistants. It stores code, documentation, notes, and behavioral rules so an assistant can recover context across sessions instead of repeatedly inspecting a repository from scratch. Indexed content is kept in Qdrant, while project awareness is based on Git repository detection and registered project scopes.

The server separates project code and documentation from reference libraries, scratchpad material, and persistent rules. The supplied setup guidance describes using project collections for codebases, libraries for external documentation, scratchpad entries for reasoning and design decisions, and rules for user-defined behavioral instructions.

## How it works

A background process watches registered project files and processes changes for ingestion. Code is divided with tree-sitter semantic chunking, with LSP integration used for active projects. Search combines dense semantic similarity with keyword matching through Reciprocal Rank Fusion, while exact substring and regular-expression queries use FTS5.

The MCP interface runs as a local stdio command named `workspace-qdrant-mcp`. Qdrant must be available separately; the example configuration points the server at `http://localhost:6333`. The project also provides the Rust-based `wqm` command-line tool for status and health checks, plus the `memexd` daemon for ongoing file monitoring and processing.

## Setup and configuration

ChrisGVE/workspace-qdrant-mcp MCP server requires a running Qdrant instance. A C compiler is needed the first time Tree-sitter grammars are compiled. Clang and LLVM are additionally needed when building `memexd` from source with its default graph backend; prebuilt binaries do not require that toolchain.

Installation options include a Homebrew package for macOS and Linux, downloadable binaries for macOS, Linux, and Windows, or a source build using the repository’s installation script. The MCP configuration passes `QDRANT_URL` in the server environment. After installation, `wqm --version` and `wqm status health` provide basic verification.

For Claude Desktop, it provides a `claude_desktop_config.json` entry using the `workspace-qdrant-mcp` command. Projects should be registered before relying on automatic watching and ingestion.

## Tools and capabilities

The six listed MCP tools are:

- `search` combines semantic and keyword retrieval across indexed content.
- `retrieve` fetches a document by identifier or metadata filter.
- `rules` lists and manages persistent behavioral rules.
- `store` saves content, registers projects, and records notes.
- `grep` searches indexed content with exact substrings or regular expressions through FTS5.
- `list` shows project files and folder structure.

## Limitations and notes

ChrisGVE/workspace-qdrant-mcp MCP server is marked alpha. Treat the current implementation as active development software and consult the repository documentation for parameters and platform-specific installation details. The server depends on Qdrant and, for some source-build paths, local C or C++ compilation tools.

_Full upstream README: https://allmcps.com/mcp/chrisgve-workspace-qdrant-mcp/readme_