DiffContext v0.2.0 - Benchmark Suite

Added

• Repository-wide context selection
• Benchmark framework
• Recall metric
• Precision metric
• Context reduction metric
• Runtime metric

Benchmark Results

simple_repo

Recall: 100%
Precision: 100%
Reduction: 0%

medium_repo

Recall: 33.33%
Precision: 100%
Reduction: 75%

Known Limitations

• Cross-file import resolution not implemented

Next

v0.3.0 Import Resolution