Find the exact files to change
Pattern Groups Generation
Creates focused pattern groups targeting specific functionality areas. Each group uses precise path patterns and content patterns with targeted exclusions to find exactly what's needed.
AI Relevance Assessment
Files are scored by actual content analysis, not just filenames or keyword matches. Quality over quantity - be conservative and selective with file inclusion for cost efficiency.
Cost-Optimized Selection
Focuses on files that will need direct modification (typically 3-10 files). Each extra file increases inference cost, so it favors brevity while safeguarding completeness.
Priority-Based Ordering
Returns files ordered by implementation priority - highest-impact files first (entry points, shared data models, core logic), then adds paths only when essential for completeness.
Frequently Asked Questions
How is this different from grep or search?
File Finder uses AI to understand your task context and scores files by actual relevance, not just text matches. It finds implementation-critical files you might miss with traditional search.
Does it work with large codebases?
Yes, it's specifically designed for large codebases where manual file discovery becomes impossible. The AI assessment scales efficiently across thousands of files.
What if it misses important files?
File Finder can expand its search scope dynamically and includes dependency analysis. You can also refine search patterns based on initial results.