Skill Seekers v3.0.0: The Universal Data Preprocessor for AI Systems

Source: Dev.to

Overview

Skill Seekers v3.0.0 is a universal documentation pre‑processor that turns any docs source into structured knowledge ready for AI systems. One command can produce 16 output formats, generate CI/CD‑ready packages, and upload results to cloud storage.

Installation

pip install skill-seekers langchain-chroma langchain-openai

Quick start

Scrape a documentation set with a single command:

skill-seekers scrape --config configs/react.json

Output formats

Using the generated documents in Python

from skill_seekers.cli.adaptors import get_adaptor
from langchain_chroma import Chroma
from langchain_openai import OpenAIEmbeddings, ChatOpenAI
from langchain.chains import RetrievalQA

# Load the LangChain adaptor and documents
adaptor = get_adaptor('langchain')
documents = adaptor.load_documents("output/react/")

# Create a vector store
vectorstore = Chroma.from_documents(documents, OpenAIEmbeddings())

# Build a RetrievalQA chain
qa_chain = RetrievalQA.from_chain_type(
    llm=ChatOpenAI(),
    retriever=vectorstore.as_retriever()
)

# Query the knowledge base
result = qa_chain.invoke({"query": "What are React Hooks?"})
print(result["result"])

The same approach works with LlamaIndex, FAISS, Qdrant, Weaviate, etc.

Cloud storage upload

# AWS S3
skill-seekers cloud upload output/ --provider s3 --bucket my-bucket

# Google Cloud Storage
skill-seekers cloud upload output/ --provider gcs --bucket my-bucket

# Azure Blob Storage
skill-seekers cloud upload output/ --provider azure --container my-container

CI/CD integration

Add the official GitHub Action to keep your AI knowledge up‑to‑date:

# .github/workflows/update-docs.yml
- uses: skill-seekers/action@v1
  with:
    config: configs/react.json
    format: langchain

Feature summary

• One‑command conversion to any of 16 formats (LangChain, LlamaIndex, Markdown, Claude, etc.)
• 26 MCP tools for AI agents (scraping, packaging, cloud upload, vector‑DB export)
• CI/CD ready with a dedicated GitHub Action
• Cloud storage support: S3, GCS, Azure
• Extensive test suite: 1,852 passing tests across 100 files, 58 k lines of Python code
• Multi‑platform: Ubuntu, macOS, Python 3.10‑3.13

Full workflow

1. Install the package (see above).
2. Scrape the target documentation: skill-seekers scrape --format langchain --config configs/react.json.
3. Load the generated documents with the appropriate adaptor.
4. Create a vector store (e.g., Chroma) and build your RAG pipeline.
5. Optional: upload the output to cloud storage or publish as a Claude skill.

All steps can be completed in about 15 minutes from raw docs to a working RAG pipeline.

Resources

• Website:
• GitHub repository:
• Documentation:
• PyPI:

Skill Seekers v3.0.0 was released on February 10, 2026.