cover-agent - Ensure Clarity in Code Testing with Automated Solutions

Cover-Agent Project Introduction

Cover-Agent is an innovative project designed to enhance code quality by streamlining the process of test generation, with a focus on unit tests. It leverages the power of Generative AI models to automatically create tests, thus increasing code coverage and ensuring a smoother development process for software engineers. This tool is a part of a larger suite aimed at automating and improving the quality of software development practices.

Overview

The Cover-Agent project consists of several key components that work together to automate the test generation process:

Test Runner: Executes commands or scripts designed to run the existing test suite, generating code coverage reports in the process.
Coverage Parser: Ensures that each new test contributes to an increase in overall code coverage, validating the effectiveness of the added tests.
Prompt Builder: Collects necessary data from the codebase to construct prompts sent to a Large Language Model (LLM). This is crucial for generating relevant and effective tests.
AI Caller: Communicates with the LLM to create tests based on the prompts provided, further integrating intelligent automation into the testing workflow.

Installation and Usage

Requirements

To start using Cover-Agent, there are a few prerequisites:

Environment Variable: You need to set the OPENAI_API_KEY in your environment to call the OpenAI API.
Code Coverage Tool: Ensure you have a tool like pytest-cov for generating Cobertura XML code coverage reports.

Additionally, if you're planning to run the tool from the source repository, you will need:

Python and Poetry: Python must be installed on your system, along with Poetry, a dependency manager for Python projects.

Standalone and Pip Installation

Cover-Agent offers flexibility depending on your setup:

Python Pip Package: Install the Cover-Agent as a Python package using the command:
```
pip install git+https://github.com/Codium-ai/cover-agent.git
```
Standalone Binary: You can also run Cover-Agent as a standalone executable, even in environments without Python, such as Docker containers. Downloads are available on the project’s release page.

Running Cover-Agent

After installation, you can run Cover-Agent using the command line for generating and validating unit tests. Here’s a basic example of how to use it:

cover-agent \
  --source-file-path "<path_to_source_file>" \
  --test-file-path "<path_to_test_file>" \
  --code-coverage-report-path "<path_to_coverage_report>" \
  --test-command "<test_command_to_run>" \
  --test-command-dir "<directory_to_run_test_command>" \
  --coverage-type "<type_of_coverage_report>" \
  --desired-coverage <desired_coverage_between_0_and_100> \
  --max-iterations <max_number_of_llm_iterations> \
  --included-files "<optional_list_of_files_to_include>"

This modular setup allows for detailed customization based on your project needs and desired outcomes.

Development Updates

Cover-Agent continually evolves with significant updates aimed at enhancing functionality and developer experience. Recent updates include:

Enhanced Database Usage: New capabilities for logging test results.
Improved Documentation: More comprehensive user examples and configuration templates.
Multi-Language Support: Docker configurations for languages like C#, TypeScript, and React, providing versatile testing environments.

Conclusion

Cover-Agent represents a major step forward in automated test generation and quality assurance in software development. By continuously expanding its features and improving usability, the Cover-Agent project invites collaboration and community engagement to refine its capabilities and explore new advancements in test automation.