fix(min-chunk): remove minsize for chunk

[icecrasher321]

Summary

OpenAI embeddings API has no min chunk size character requirements. So this PR removes it from here too setting default to 1 char.

Type of Change

• Bug fix

Testing

Tested manually creating doc with <10 chars.

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name	Status	Preview	Comments	Updated (UTC)
sim	✅ Ready (Inspect)	Visit Preview	💬 Add feedback	Aug 8, 2025 7:41pm

1 Skipped Deployment

Name	Status	Preview	Comments	Updated (UTC)
docs	⬜️ Skipped (Inspect)			Aug 8, 2025 7:41pm

[greptile-apps]

Greptile Summary

This PR systematically reduces the minimum chunk size constraint across Sim's document processing and knowledge base system from 50-100 characters to 1 character. The changes span the entire document processing pipeline, including:

• Frontend validation: Upload modals and create modals now allow chunks as small as 1 character
• API validation: Knowledge base creation and document processing APIs updated to accept minimum chunk sizes of 1
• Core processing: The TextChunker class, DocsChunker, and document processor functions all now default to a minimum chunk size of 1
• Tool integration: Knowledge tools updated to process documents with 1-character minimum chunks

The changes maintain backward compatibility by preserving existing parameter structures while changing default values. This modification affects how documents are chunked for RAG (Retrieval Augmented Generation) applications, allowing the system to preserve very small text fragments like single words, abbreviations, code snippets, or structured data elements that were previously filtered out.

The chunking system uses hierarchical text splitting that attempts to respect semantic boundaries (sentences, paragraphs, etc.) but falls back to character-level splitting when necessary. With this change, even single-character chunks will be preserved if larger semantic chunks cannot be formed, ensuring maximum content retention during document processing.

PR Description Notes:

• The PR description is incomplete - all sections remain as template placeholders without actual content describing the changes or testing approach

Confidence score: 2/5

• This PR introduces significant risk due to potential performance and quality impacts from creating many small chunks
• Score reflects concerns about storage costs, retrieval quality degradation, and inconsistent implementation across the codebase
• Pay close attention to chunking logic files and API validation schemas for potential inconsistencies

_{11 files reviewed, 1 comment}

_{Edit Code Review Bot Settings | Greptile}