Back to servers

NOUZ MCP Server

MCP Server for local knowledge management. Semantic + keywords + tags

glama

Updated

Apr 4, 2026

Claim this server

NOUZ — MCP Server for local knowledge management

v2.1.0

Install

pip install -r requirements.txt

Run

export OBSIDIAN_ROOT=./obsidian
export EMBED_PROVIDER=ollama  # openai, gigachat
python server.py

Config

Edit config.yaml:

Field	Type	Description
`mode`	string	luca / prizma / sloi
`levels`	dict	core=1, pattern=2, module=3, quant=4, artifact=5
`thresholds.semantic_bridge_threshold`	float	0.55 default
`etalons`	list	Core signs with text for embedding
`prizma_modes`	dict	Keywords per thinking style

Tools

Basic

read_file(path) — read note with YAML frontmatter
write_file(path, content, metadata) — write note with YAML
list_files(level, sign, tags, subfolder) — filter files
index_all(with_embeddings) — index to DB
embed(text) — get embedding

Navigation

get_parents(path) — files linking to this file
get_children(path) — files this file links to

Semantics (prizma / sloi)

calibrate_cores() — recalculate etalon embeddings
recalc_signs() — recalculate sign_auto
recalc_core_mix() — recalculate core_mix
suggest_metadata(path) — suggest level/sign
suggest_parents(path) — suggest links by embeddings
format_entity_compact(path) — entity formula

Modes

Mode	Level Strict	Semantics	Description
luca	❌	❌	Simple graph
prizma	❌	✅	Semantic bridges + core_mix
sloi	✅	✅	Strict 5-level hierarchy

File Format

---
level: 2
sign: Ψ
tags: [systems]
---
# Content here

Env Vars

Variable	Default	Description
`OBSIDIAN_ROOT`	./obsidian	Vault path
`EMBED_PROVIDER`	openai	openai / gigachat / ollama
`EMBED_API_URL`	http://127.0.0.1:1234/v1	API endpoint
`EMBED_MODEL`	—	Model name
`EMBED_API_KEY`	—	API key

Copyright (c) 2026 KVANTRA. MIT License.

Reviews

No reviews yet

Sign in to write a review