eval

Automated code quality evaluation for SDD workflows is achieved by collecting git branch changes and applying third-party assessment models. The process detects the target directory, gathers feature-wide changes, runs the evaluation, and outputs a detailed report.

Code change collection works by automatically detecting the main code directory in your current SDD branch and gathering feature-wide changes. It then generates a config.result.json file inside the changes/{FEATURE_DIR}/evalset directory.

Yes, automated code evaluation requires a Python runtime with standard packages and access to a third-party evaluation model endpoint. These prerequisites allow the script to gather changes, run assessments, and output results.

The best way to output code evaluation results is through automated console logging and file generation. The evaluation prints progress to the console and writes a detailed result.txt file directly into the changes/{FEATURE_DIR}/evalset directory.

Yes, you can use third-party quality assessment models to evaluate git code changes. The automation workflow connects to your evaluation model endpoint, processes the collected feature-wide changes, and generates a detailed quality report.

Detecting the main code directory automatically ensures that feature-wide changes are gathered accurately from the correct SDD branch path. This prevents manual path configuration errors and guarantees that the evaluation model assesses the intended codebase modifications.