skills.rest
latestaiagents
latestaiagentsOFFICIAL
Explore all skills by this author →
GitHub

chunking-strategies

Guide document chunking strategies for RAG systems with code examples.

5|1|Updated Feb 4, 2026
Search Tags:#document processing#embeddings#rag#retrieval#chunking#text splitting
One-click install
npx skills add https://github.com/latestaiagents/agent-skills --skill chunking-strategies-latestaiagents
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: chunking-strategies
Source: https://github.com/latestaiagents/agent-skills/tree/main/plugins/rag-architect/skills/chunking-strategies
Command: npx skills add https://github.com/latestaiagents/agent-skills --skill chunking-strategies-latestaiagents

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the challenge of effectively splitting documents into meaningful chunks for retrieval-augmented generation (RAG) systems, ensuring better retrieval accuracy and context preservation.

Core Features & Use Cases

  • Diverse Chunking Methods: Supports fixed-size, recursive, semantic, and document-specific (Markdown, Code, HTML) chunking.
  • Optimized Chunk Sizing: Provides guidelines for token counts and overlap based on content type.
  • Advanced Techniques: Includes implementation details for parent-child chunking.
  • Metadata Enrichment: Demonstrates how to add crucial metadata to chunks for better filtering and context.
  • Use Case: When building a RAG system for technical documentation, use this skill to select the most appropriate chunking strategy (e.g., recursive for prose, language-specific for code) to maximize the relevance of retrieved information.

Quick Start

Use the chunking-strategies skill to apply recursive character splitting with a chunk size of 1000 tokens and an overlap of 200 tokens to the provided document.

Frequently Asked Questions about chunking-strategies

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is the best way to chunk documents for a RAG system?▼

The best chunking strategy for RAG depends on content type, utilizing fixed-size, recursive, semantic, or document-specific splitting to optimize retrieval accuracy and context preservation.

How do I split mixed content types like code and Markdown for retrieval?▼

To split mixed content types for retrieval, apply document-specific methods, using recursive splitting for prose and language-specific rules for code to maintain structural integrity.

How much overlap should I use when chunking text for embeddings?▼

Overlap for text chunking in embeddings should be adjusted based on content type, with specific token count guidelines provided to balance context preservation against retrieval redundancy.

When should I use semantic chunking instead of fixed-size text splitting?▼

Use semantic chunking over fixed-size text splitting when meaning and context boundaries are critical, whereas fixed-size splitting suits uniform prose where consistent token counts matter.

How do I add metadata to chunks to improve RAG retrieval quality?▼

Metadata enrichment improves RAG retrieval quality by attaching crucial contextual information to chunks, enabling better filtering and more accurate context preservation during retrieval.

What is parent-child chunking and when should I use it?▼

Parent-child chunking links smaller retrieved chunks to larger parent chunks, improving retrieval accuracy by providing broader context without sacrificing search precision.

Related Skills

Explore More Software Engineering Skills→
javierhbrjavierhbr

docling

Parse PDF, DOCX, PPTX, HTML, and images into Markdown, HTML, or JSON.

Community
Advanced
christian-bromannchristian-bromann
3

langchain-embeddings-vectorstores

Convert text into embeddings and index them into vector stores for semantic search.

Community
Advanced
zeronx798zeronx798

chroma

Manage vector collections and perform similarity searches with metadata filtering.

Community
Intermediate
GongLingRuiGongLingRui
330

text-splitter

Split long text into uniform chunks at sentence or paragraph boundaries.

Community
Intermediate