Pageindex RAG — AI skill for Claude Code

PageIndex RAG Skill

Reasoning-based retrieval architecture replacing vector databases with hierarchical navigation

Overview

PageIndex replaces vector-based similarity search with LLM-driven hierarchical navigation, achieving **98.7% accuracy** on financial document benchmarks by reasoning through document structure instead of matching embeddings.

Installation

npx skills add mmtmr/pageindex-rag -g -y

What This Skill Provides

• RAG Architecture Design: Hierarchical table-of-contents indices with LLM-driven navigation
• Implementation Patterns: Practical code examples for building PageIndex systems
• Vector RAG Migration: Converting existing vector-based systems to reasoning-based retrieval
• Document Indexing: Strategies for structured documents (financial reports, legal contracts, technical manuals)

Use Cases

Use this skill when:

1. Implementing RAG for long structured documents (financial reports, legal contracts, technical manuals)
2. Improving existing vector-based RAG systems with poor accuracy on structured content
3. Designing document indexing strategies with semantic coherence
4. Explaining PageIndex concepts including reasoning-based retrieval, hierarchical navigation, and cross-reference following
5. Handling documents with internal references and multi-turn conversations

Core Innovation

**Why Vector RAG Fails:**

• Query-knowledge mismatch (surface semantics ≠ task relevance)
• Hard chunking fragments contextual continuity
• Context window deterioration with 10-20 chunks
• Cannot follow cross-references

**PageIndex Solution:** Replace vector databases with hierarchical tree indices stored as JSON, enabling LLM-driven navigation through document structure.

Key Features

• 98.7% accuracy on financial document benchmarks
• Hierarchical navigation with semantic coherence
• Cross-reference following capability
• Multi-turn conversation support
• No embedding dependency

References

• [Architecture