Installation
This guide covers how to install AgenticAssure and verify that everything is working.
System Requirements
- Python 3.10 or higher. AgenticAssure uses modern Python features including
X | Yunion types and other 3.10+ syntax. Python 3.10, 3.11, and 3.12 are all supported. - pip (included with Python) or any PEP 517-compatible installer.
- Operating system: Windows, macOS, or Linux. AgenticAssure is a pure-Python package with no platform-specific dependencies.
Basic Installation
Install the core package from PyPI:
pip install agenticassureThis installs AgenticAssure and its core dependencies:
| Dependency | Version | Purpose |
|---|---|---|
| pydantic | >= 2.0 | Data validation and models |
| pyyaml | >= 6.0 | YAML scenario file parsing |
| click | >= 8.0 | CLI framework |
| rich | >= 13.0 | Formatted terminal output |
| jsonschema | >= 4.0 | Scenario file schema validation |
Installing with Extras
AgenticAssure provides optional dependency groups for additional functionality. Install them using the bracket syntax:
Similarity Scorer
Enables the similarity scorer, which uses sentence-transformers to compute semantic similarity between expected and actual outputs. Useful when you care about meaning rather than exact wording.
pip install agenticassure[similarity]This installs sentence-transformers>=2.0, which in turn pulls in PyTorch and Hugging Face libraries. The first run will download a sentence-transformer model (typically ~100MB).
OpenAI Adapter
Installs the OpenAI Python SDK for use with the built-in OpenAI adapter:
pip install agenticassure[openai]LangChain Adapter
Installs LangChain for use with the built-in LangChain adapter:
pip install agenticassure[langchain]All Optional Dependencies
Install everything at once:
pip install agenticassure[all]This is equivalent to installing similarity, openai, and langchain extras together.
Development Dependencies
For contributing to AgenticAssure or running its test suite:
pip install agenticassure[dev]This installs pytest, pytest-cov, ruff, and black.
Installing from Source
To install from the GitHub repository directly:
pip install git+https://github.com/agenticassure/agenticassure.gitDevelopment Mode
If you want to work on AgenticAssure itself, clone the repository and install in editable mode:
git clone https://github.com/agenticassure/agenticassure.git
cd agenticassure/core
pip install -e ".[dev]"This installs the package in editable mode so that changes to the source code take effect immediately without reinstalling.
Virtual Environments
It is strongly recommended to use a virtual environment to avoid dependency conflicts with other Python projects:
# Create a virtual environment
python -m venv .venv
# Activate it (Linux/macOS)
source .venv/bin/activate
# Activate it (Windows)
.venv\Scripts\activate
# Then install
pip install agenticassureVerifying the Installation
After installing, verify that the CLI is available and working:
agenticassure --versionExpected output:
agenticassure, version 0.3.0You can also verify the Python import:
import agenticassure
print(agenticassure.__version__)
# 0.3.0Common Installation Issues
command not found: agenticassure
The CLI entry point was not added to your PATH. This usually means:
- You installed into a virtual environment that is not activated. Activate it with
source .venv/bin/activate(Linux/macOS) or.venv\Scripts\activate(Windows). - You installed with
--userand your user-levelbindirectory is not on yourPATH. On Linux/macOS, add~/.local/binto yourPATH. On Windows, the user scripts directory is typically%APPDATA%\Python\PythonXX\Scripts.
ModuleNotFoundError: No module named 'agenticassure'
This typically means you are running Python from a different environment than the one where you installed the package. Make sure you have activated the correct virtual environment.
ERROR: No matching distribution found for agenticassure
Check that you are using Python 3.10 or higher:
python --versionIf you have multiple Python versions installed, you may need to use python3 or python3.12 explicitly:
python3 -m pip install agenticassureSentence-transformers installation is slow or fails
The similarity extra pulls in PyTorch, which is a large download. If the install hangs or fails:
- Make sure you have sufficient disk space (PyTorch can be 1-2GB).
- On systems without a GPU, pip will install the CPU-only variant automatically.
- If you are behind a corporate proxy, ensure
pipis configured to use it.
Permission errors on install
Avoid using sudo pip install. Instead, use a virtual environment (recommended) or install with the --user flag:
pip install --user agenticassureNext Steps
With AgenticAssure installed, head to the Quickstart to run your first test suite.