Automated code generation.
Important
This is a template repository. To use its functionalities, use this template to create own repository.
The system is designed to enable connecting different kinds of Large Language Models. To successfully connect an LLM, the following actions need to be taken:
- The file
.sidmd/generate_code.pyshould be updated with implementation-specific connection and request details. - If the LLM usage requires installing specific Python libraries, they should be listed in
.sidmd/install_dependencies.sh. - For security reasons, if an API token is necessary, it should be defined under a name
AI_API_KEYin repository's secrets (Settings>Secrets and variabless>Actions>Secrets).
- Start by choosing a suitable issue template:
Initiation— for creating initial file structure.Feature— for implementing a new feature.Bug— for fixing a bug.
- Wait for the AI to finish generating code. This will be confirmed by a comment under the issue, together with a link to a branch where the change was made.
- Follow the link to review changes made on a branch. If any problems arises or there is a need to further modify the code, indicate the expected changes by a comment under the original issue. Repeat the process until the results are satisfactory.
- When the result meets expectations, close the issue. This will trigger an automatic creation of a pull request to the main branch.