Configure code chunking in GrepAI. Use this skill to optimize how code is split for embedding.

GrepAI Chunking Configuration This skill covers how GrepAI splits code files into chunks for embedding, and how to optimize chunking for your codebase. When to Use This Skill Optimizing search accuracy Adjusting for code style (verbose vs. concise) Troubleshooting search results Understanding how indexing works What is Chunking? Chunking is the process of splitting source files into smaller segments for embedding: Why Chunking Matters Embedding models have optimal input sizes: Too large chunks: Less precise search results Too small chunks: Lost context, fragmented results Just right: Good balance of precision and context Configuration Basic Settings Understanding Parameters Chunk Size The target number of tokens per chunk. Size Effect 256 More precise, less context 512 Balanced (default) 1024 More context, less precise Overlap Tokens shared between adjacent chunks. Preserves context at boundaries. Overlap Effect 0 No overlap, may lose context at boundaries 50 Standard overlap (default) 100 More context, larger index Visualization With size=512 and overlap=50: Recommended Settings by Language Verbose Languages (Java, C ) Concise Languages (Go, Python) Very Concise (Rust, Zig) Recommended Settings by Codebase Small Functions (Microservices) Large Classes (Monolith) Mixed Codebase How Tokens are Counted GrepAI uses approximate token counting: ~4 characters = 1 token (for English text) Code varies based on identifiers and syntax Example: ≈ 45 tokens Impact on Index Size Larger overlap = more chunks = larger index: Size Overlap Chunks per 10K tokens Index Impact 512 0 ~20 Smallest 512 50 ~22 Standard 512 100 ~24 +10% 256 50 ~44 +100% Impact on Search Quality Too Small Chunks (size: 128) Just Right (size: 512) Too Large Chunks (size: 2048) Experimentation Testing Different Settings 1. Try smaller chunks for more precise results: 2. Re index: 3. Test with searches: 4. Adjust and repeat until satisfied. Comparing Results Before changing settings, save a search result: After changing settings and re indexing: Chunk Boundaries GrepAI tries to split at logical boundaries: 1. Empty lines (function/class boundaries) 2. Closing braces 3. Statement ends This means actual chunk sizes may vary slightly from the target. Best Practices 1. Start with defaults: 512/50 works well for most codebases 2. Adjust based on code style: Verbose = larger, concise = smaller 3. Test with real queries: See what your searches return 4. Re index after changes: Must regenerate embeddings 5. Consider overlap: Don't set to 0 unless index size is critical Common Issues ❌ Problem: Search results are too fragmented ✅ Solution: Increase chunk size: ❌ Problem: Search results have too much irrelevant context ✅ Solution: Decrease chunk size: ❌ Problem: Results miss related code at function boundaries ✅ Solution: Increase overlap: ❌ Problem: Index is too large ✅ Solutions: Decrease overlap Increase chunk size Add more ignore patterns Output Format Chunking status:

grepai-chunking