---
id: nvidia-nat-atif
version: "1.8.0"
license: Apache-2.0
license_treatment: permissive
maintenance: active
---

# nvidia-nat-atif — Subpackage for ATIF schema models in NVIDIA NeMo Agent Toolkit

License: permissive · Maintenance: active · Downloads: 173.9K/mo

## What it is and what it does

This is a subpackage of the NVIDIA NeMo Agent Toolkit that provides ATIF schema models for structuring and validating agent data. It leverages pydantic for data validation, enabling developers to work with standardized, type-safe representations of agent interactions and evaluation metrics.

The package targets Python 3.11–3.13 and is maintained as part of an active, well-starred GitHub repository. It is intended for teams building or evaluating AI agents within the NeMo ecosystem, where consistent schema enforcement across agent interactions and evaluation pipelines is needed.

Use it for:

- Validate and structure agent interaction logs using standardized ATIF schemas within NeMo Agent Toolkit workflows
- Build evaluation pipelines that consume and process agent behavior data in a type-safe, schema-compliant format
- Integrate agent evaluation metrics and results into larger NeMo-based AI systems with consistent data models
- Develop custom agent evaluation tools that rely on pydantic-validated ATIF data structures

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides ATIF schema models for the NVIDIA NeMo Agent Toolkit, enabling structured representation and validation of agent evaluation and interaction data.

Yes, if you are building or evaluating agents within the NVIDIA NeMo ecosystem and need standardized schema models for agent interactions and evaluation. The low install friction, active maintenance, permissive license, and zero known vulnerabilities make it a safe choice. If you are not working with NeMo agents or do not need structured ATIF schemas, it is not applicable.

## Install

pip install nvidia-nat-atif
uv add nvidia-nat-atif
poetry add nvidia-nat-atif

## Installing nvidia-nat-atif

Before you install: Low install friction with a single runtime dependency on pydantic. The package is actively maintained with recent commits and sits within the top 15000 PyPI packages by download volume.

License in practice: Licensed under Apache-2.0 (permissive), allowing free use, modification, and distribution with minimal restrictions—suitable for both open and commercial projects.

Quickstart:

pip install nvidia-nat-atif

from nvidia_nat_atif import <schema_model>
import pydantic

# Use pydantic-validated ATIF schema models for agent data

Requires Python 3.11 or later (supports 3.11, 3.12, 3.13)

Verify before relying:

- What specific ATIF schema models are included and their intended use cases within agent evaluation
- Whether this subpackage can be used independently or requires other NeMo Agent Toolkit components
- Performance characteristics or scalability limits for large-scale agent evaluation workflows

## Package facts

- License: Apache-2.0 (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: active
- Downloads: 173.9K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

atif schema models, nemo agent toolkit, agent evaluation framework, structured agent data, nvidia nemo agents, agent interaction schemas, nemo toolkit subpackage, agent-toolkit, schema-validation, nvidia-nemo

[View on SkillFed](https://skillfed.io/packages/nvidia-nat-atif) · [View on PyPI](https://pypi.org/project/nvidia-nat-atif/)