v0.8.0
Changelog
[Version 0.8.0] - 2023-10-10
This release introduces significant organizational changes to the git-gpt
project, enhances its functionality, and improves usability through a modular command structure. We've also updated several command options for clarity and specificity.
Added
- Introduced a new command
ask_command.py
to enable users to ask custom questions about code diffs with specified models and commit ranges. - Added
changelog_command.py
for generating changelogs based on recent commits using the OpenAI API. - Integrated
commit_command.py
to facilitate crafting of commit messages based on diff analysis. - Included
config_command.py
to assist users in configuring the settings like API keys and token limits. - Added
issue_command.py
for generating GitHub-style development issues from code diffs. - The
quality_command.py
was introduced for executing quality checks on code changes, focusing on adherence to coding standards, potential bugs, and security vulnerabilities.
Changed
- Revised the README to include a detailed project structure, improving understanding of the codebase organization.
- Refactored the command-line interface in
main.py
to modularize individual commands, enhancing maintainability and scalability. - Adjusted token option names for clarity, such as
--max-tokens-quality
,--max-tokens-issue
, and--max-tokens-changelog
to specify module-specific token limits.
Removed
- Outdated and redundant amalgamated command implementations in
main.py
, which have now been separated into dedicated command files for better organization.
Fixed
- Corrected and improved several command flags for consistency in usage, including
--max-tokens
override options for specific prompts.
Security
- Enhanced the security of the codebase by examining diff outputs for vulnerabilities in the
quality_command.py
.
The development team has revised several core functionalities to adhere to our strategic goal of enhancing automation and configurability for developer workflows. This version marks a leap towards more organized and user-friendly interfaces.