Skool RAG Pipeline
Goal
Query Skool community content using a RAG (Retrieval-Augmented Generation) pipeline with vector search and reranking.
Scripts
./scripts/skool_rag_prepare.py- Prepare content for indexing./scripts/skool_rag_index.py- Index content in Pinecone./scripts/skool_rag_query.py- Query the knowledge base
Pipeline
1. Prepare Content
python3 ./scripts/skool_rag_prepare.py --community makerschoolScrapes and chunks community content.
2. Index in Pinecone
python3 ./scripts/skool_rag_index.py --input .tmp/skool_chunks.jsonCreates OpenAI embeddings and stores in Pinecone.
3. Query
python3 ./scripts/skool_rag_query.py --query "How do I get my first client?"Pipeline:
- OpenAI embeddings for query
- Pinecone vector search
- Cohere reranking
- Claude response generation
Environment
PINECONE_API_KEY=your_key
OPENAI_API_KEY=your_key
COHERE_API_KEY=your_key
ANTHROPIC_API_KEY=your_keySchema
Inputs
| Name | Type | Required | Description |
|---|---|---|---|
query | string | Yes | Natural language question to search for |
community | string | No | Community slug to index (default: makerschool) |
Outputs
| Name | Type | Description |
|---|---|---|
answer | string | AI-generated answer with source references |
Credentials
| Name | Source |
|---|---|
PINECONE_API_KEY | .env |
OPENAI_API_KEY | .env |
COHERE_API_KEY | .env |
ANTHROPIC_API_KEY | .env |
Composable With
Skills that chain well with this one: skool-monitor
Cost
Pinecone + OpenAI embeddings + Cohere reranking + Claude