skills.rest
mindmorass
mindmorassCOMMUNITY
Explore all skills by this author →
GitHub

knowledge-ingestion-patterns

Ingest PDFs, websites, code repos, and notes into vector databases with chunking and metadata schemas.

2|Updated Jan 2, 2026
Search Tags:#rag#metadata#ingestion#deduplication#chunking#vector-databases
One-click install
npx skills add https://github.com/mindmorass/reflex --skill knowledge-ingestion-patterns
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: knowledge-ingestion-patterns
Source: https://github.com/mindmorass/reflex/tree/main/plugins/reflex/skills/knowledge-ingestion-patterns
Command: npx skills add https://github.com/mindmorass/reflex --skill knowledge-ingestion-patterns

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides best practices for ingesting diverse content into vector databases and RAG systems, ensuring efficient chunking, metadata, and retrieval quality.

Core Features & Use Cases

  • Pattern-based ingestion for PDFs, websites, code repositories, and research notes to optimize chunking and metadata.
  • Rich metadata schemas and deduplication rules to improve retrieval accuracy and provenance.
  • Use Case: Ingest a mixed set of sources (PDFs, web pages, and notes) and generate consistent chunks with metadata ready for indexing.

Quick Start

Ingest a PDF, a website, or notes by providing the content source and desired chunk size; the skill will outline the ingestion plan and metadata schema.

Frequently Asked Questions about knowledge-ingestion-patterns

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

FAQPage Schema
How do I chunk and ingest PDF content into a vector database?▼

Chunking breaks PDFs into appropriately sized segments optimized for vector retrieval. This Skill applies pattern-based ingestion rules to PDFs, extracting content with metadata schemas that preserve context and improve retrieval accuracy when indexed into vector databases.

What's the best way to deduplicate content when ingesting from multiple sources?▼

Deduplication removes redundant data across PDFs, websites, code repositories, and notes before indexing. This Skill prescribes deduplication controls and metadata rules to ensure consistent, high-quality chunks and prevent duplicate retrieval results in RAG systems.

Can I ingest diverse content types—PDFs, websites, and code repos—into the same RAG system?▼

Yes. This Skill handles mixed-source ingestion across PDFs, websites, code repositories, and research notes using pattern-based rules tailored to each format. It generates consistent chunks with unified metadata schemas ready for vector indexing across all source types.

How do I structure metadata to improve vector retrieval quality?▼

Rich metadata schemas tag chunks with provenance, context, and source information, enabling more precise retrieval. This Skill prescribes metadata structures that work across diverse content types to ensure RAG systems return accurate, contextual results.

What metadata should I include when indexing content for RAG systems?▼

Metadata includes source type, chunk boundaries, provenance details, and contextual tags that help vector databases retrieve the most relevant chunks. This Skill defines schemas that work across PDFs, websites, code, and notes to optimize retrieval consistency.

Related Skills

Explore More Data & Analytics Skills→
maxvaegamaxvaega
146

pdf-extractor

Extract text and metadata from PDF documents via Python scripts.

Community
Intermediate
rukikotoorukikotoo
6

rag

Query the econometrics RAG knowledge base to retrieve relevant document chunks.

Community
Intermediate
ai-kitchen-incai-kitchen-inc
2

data-layer

Chunk documents, generate embeddings, and index vectors in Pinecone for RAG.

Community
Intermediate
tankpkgtankpkg
1

@tank/rag-production

Design production RAG pipelines with chunking, retrieval, reranking, and context assembly.

Official
Advanced