repowiserepowise

App

nvidia / auto-ontology

Auto Ontology is a Python framework for building semantic, graph-based reasoning layers over data. It enables deterministic query generation, better context retrieval, and more explainable AI for LLM and agent workflows.

★ 12,748stars on GitHubPythonApache-2.0
Install auto-ontologyView on GitHub★ 12.7kSee this for your own repoAsk this codebase

Repowise maps public repos from their source. Not affiliated with nvidia. What is Repowise?

github.com/nvidia/auto-ontologyIndexed at 22e22c8

What is auto-ontology?

auto-ontology is an open library that builds a context layer from enterprise metadata so AI agents know where data lives, what it means and whether to trust it. It generates an ontology of business terms, SQL attributes and table relationships from connected databases, and people certify that ontology before agents rely on it. The core code is written in Python, with a Next.js web frontend. To install it, run git clone https://github.com/nvidia/auto-ontology.git and follow the steps below. Its code health is 7.0 out of 10 (Fair), ranked #122 of 4,000+ indexed repos. The last commit to auto-ontology was on Nov 13, 2026, and 13 people have contributed.

As of commit 22e22c8, indexed Nov 14, 2026. What it is comes from a model reading the README; the rest is measured from the source and git history, not from the maintainer.

For agentsauto-ontology.md

Repo facts: type, license, health, rank, contributors
auto-ontology facts
TypeApp
LicenseApache-2.0
Code health7.0 / 10 (Fair)
Health rank#122 of 4,000+ indexed repos
Last commitNov 13, 2026
Contributors13
Bus factor2
Size995 files, 134,289 lines

Install auto-ontology

Commands copied from the README

  1. 1Clone the auto-ontology repository.
    git clone https://github.com/nvidia/auto-ontology.git
  2. 2Move into the auto-ontology folder.
    cd auto-ontology
  3. 3Copy the example environment file to .env, then set the API key, passwords and admin credentials in it.
    cp .env.example .env
  4. 4Start Postgres.
    docker compose up -d --wait postgres
  5. 5Build and start the rest of the stack, then sign in with the admin credentials from .env.
    docker compose up -d --build

Full install notes on GitHub

Explore auto-ontology

Ask about this code in your agent

This connects Repowise to auto-ontology's code. It does not install auto-ontology. To install auto-ontology, use the install steps.

Useful if you want to change or fork auto-ontology: your agent can ask how it is put together and get answers that cite its files.

Show the connect command (Claude Code, Cursor, Codex, VS Code)

Run it in a terminal, then type /mcp in Claude Code and sign in with Repowise. No key to paste.

Docs
bash
claude mcp add --transport http repowise https://api.repowise.dev/mcp/nvidia/auto-ontology
Claude Code, Claude, Cursor and VS Code sign you in on first use, free. Codex and Windsurf need an API key: sign in to create one.Sign in for a key

This repo is public, so anyone with a free Repowise account can connect.

Try asking

  • get_risk(["auto_ontology/retrieval/text_to_sql/prompts.py"])
  • get_answer("how is auto-ontology put together?")
  • get_why("why is auto-ontology structured this way?")

Ask about auto-ontology

Answers come from auto-ontology's README and source, with the files they cite. Two free answers a day without an account.

Featured answer

Give me an overview of auto-ontology

auto-ontology is an open Python library from nvidia/auto-ontology that builds a context layer from enterprise metadata so AI agents know where data lives, what it means and whether to trust it. It generates an ontology of business terms, SQL attributes and table relationships from connected databases, and people certify that ontology before agents rely on it. The ingestion service entry point, auto_ontology/ingestion_service/main.py, extracts schemas, profiles columns and samples values into the catalog. The auto_ontology/catalog module records database structure and SQL as nodes and edges in a graph store, and auto_ontology/dal holds every query behind the catalog, glossary, tags and exploration graph. The generated ontology lives in Postgres with pgvector embeddings for retrieval, alembic manages its migrations, and the frontend directory holds the Next.js web app. To start reading, begin with the ingestion service entry point and the catalog and dal modules, since the 995 files fall mostly into the Application (449), API (131) and Utility (68) layers.

Files per layer in auto-ontology9 layers.

Basis: Inferred · knowledge graph· as of 22e22c8

Want this for your own codebase? Map your repo free

About this page

Get a page like this for your own codebase

Repowise built this map of nvidia/auto-ontology from its source and git history. It parses every symbol, draws the dependency graph, scores each file's health from complexity, duplication, coverage and churn, finds hotspots and owners in the history, and writes documentation for every module. Your coding agent reads the same index over MCP, so its answers cite real files.

Free for 2 public repos, with a page like this one. No card. See how Repowise works

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed. Star Repowise on GitHub

nvidia/auto-ontology FAQ

What is auto-ontology?

auto-ontology is an open library that builds a context layer from enterprise metadata so AI agents know where data lives, what it means and whether to trust it. It is built for agentic analytics over structured data.

How does auto-ontology work?

auto-ontology generates an ontology of terms, SQL attributes and relationships from connected databases, and people certify it. The agent then answers questions by retrieving candidate terms and columns by embedding search, writing SQL and running it.

How do I install auto-ontology?

Clone the repository, copy .env.example to .env and set the API key and passwords, then run docker compose up -d --build. Docker with Compose v2, uv and an NVIDIA API key are needed first. See the install steps

What language is auto-ontology written in?

auto-ontology is written mainly in Python. Its web frontend is a Next.js app written in TypeScript.

How is the auto-ontology code organized?

The auto_ontology package holds the catalog, the data access layer in dal and the semantic code. The frontend directory is the Next.js web app, and alembic holds the Postgres schema migrations.

Who is auto-ontology for?

auto-ontology is for data teams and developers who ground AI agents in business definitions, check generated queries and serve that context through MCP, CLI and API.

What license does auto-ontology use?

auto-ontology is licensed under Apache-2.0.

Code health

6.9out of 10Fair

Hotspots average 5.7, the place fixes pay off most.

Full health report

On the leaderboards:#32 of 64 Python repos

Key features of auto-ontology

  • Translates natural-language questions into SQL and runs them against connected databases.
  • Builds a semantic catalog and ontology layer over relational sources, stored in Postgres alongside the catalog it describes.
  • Handles authentication and role-based access control using Better Auth.
  • Provides a web UI for chat, analysis, data catalog, analytics and settings.
  • Serves ontology context to agents through MCP, CLI and API.
  • Deploys with a Helm chart for Kubernetes or with Docker Compose for a local stack.

How nvidia/auto-ontology works

Source databases are read by the ingestion service, whose entry point is auto_ontology/ingestion_service/__main__.py, which extracts schemas, profiles columns and samples values into the catalog. The auto_ontology/catalog module records each database's structure and the SQL run against it as nodes and edges in a graph store. The auto_ontology/dal module holds every database query behind the catalog, the glossary of Terms, tags and the exploration graph. The generated ontology is stored in Postgres with embeddings in pgvector for retrieval, and alembic keeps the schema migration history that moves that database from one revision to the next. The frontend directory is the Next.js web app where people chat with the data, browse the catalog and refine terms.

The codebase has 995 files and 8,271 symbols in 6 modules, led by Python, TypeScript and Dockerfile. Code health is 6.9 out of 10, rated fair.