rag-scaffold

Name: rag-scaffold
Author: floflo777

33stars3forks

How to install rag-scaffold?

floflo777/claude-rag-skills/rag-scaffold

$npx -y skills add floflo777/claude-rag-skills --skill rag-scaffold

Installs into the current project.

Summary

Generate production-ready RAG pipeline boilerplate code with best practices built-in.

Files · 1

View on GitHub

SKILL.md

1	# RAG Scaffold Skill
2
3	Generate production-ready RAG pipeline boilerplate code with best practices built-in.
4
5	## When to Use
6
7	Use `/rag-scaffold` when:
8	- Starting a new RAG project from scratch
9	- Need a reference implementation with best practices
10	- Want to quickly prototype a RAG system
11	- Learning RAG architecture patterns
12
13	## Scaffold Options
14
15	### Framework Choice
16	1. Python + LangChain - Most popular, extensive ecosystem
17	2. Python + LlamaIndex - Document-focused, great for complex pipelines
18	3. Python + Vanilla - No framework, full control
19	4. TypeScript + LangChain.js - For Node.js environments
20	5. Ailog API - Managed RAG-as-a-Service (simplest)
21
22	### Vector Store Choice
23	1. Qdrant - High performance, filtering, hybrid search
24	2. Pinecone - Managed, scalable, serverless option
25	3. ChromaDB - Simple, local-first, good for prototyping
26	4. Weaviate - GraphQL API, hybrid search
27	5. Milvus - High scale, GPU acceleration
28
29	### LLM Provider
30	1. OpenAI - GPT-4o, GPT-4o-mini
31	2. Anthropic - Claude 3.5 Sonnet, Claude 3 Opus
32	3. Mistral - Mistral Large, Mistral Small
33	4. Local - Ollama, vLLM
34
35	## How to Generate Scaffold
36
37	When the user invokes `/rag-scaffold`, ask:
38
39	1. What's your use case? (Customer support, documentation search, code assistant)
40	2. Preferred framework? (LangChain, LlamaIndex, Vanilla, Ailog API)
41	3. Vector store? (Qdrant, Pinecone, ChromaDB, etc.)
42	4. LLM provider? (OpenAI, Anthropic, Mistral)
43	5. Features needed?
44	- [ ] Hybrid search (dense + BM25)
45	- [ ] Reranking
46	- [ ] Conversation memory
47	- [ ] Streaming responses
48	- [ ] Source citations
49	- [ ] Multi-tenancy
50
51	## Scaffold Templates
52
53	### Template 1: Python + LangChain + Qdrant + OpenAI
54
55	Project Structure:
56	```
57	my-rag-project/
58	├── src/
59	│ ├── __init__.py
60	│ ├── config.py # Configuration management
61	│ ├── embeddings.py # Embedding service
62	│ ├── vectorstore.py # Vector store operations
63	│ ├── retriever.py # Retrieval logic
64	│ ├── generator.py # LLM generation
65	│ ├── rag_pipeline.py # Main RAG orchestration
66	│ └── chunker.py # Document chunking
67	├── scripts/
68	│ ├── index_documents.py # Indexing script
69	│ └── evaluate.py # Evaluation script
70	├── tests/
71	│ ├── test_retriever.py
72	│ └── test_generator.py
73	├── .env.example
74	├── requirements.txt
75	├── docker-compose.yml # Qdrant + Redis
76	└── README.md
77	```
78
79	config.py:
80	```python
81	from pydantic_settings import BaseSettings
82	from functools import lru_cache
83
84	class Settings(BaseSettings):
85	# OpenAI
86	openai_api_key: str
87	embedding_model: str = "text-embedding-3-small"
88	llm_model: str = "gpt-4o-mini"
89
90	# Qdrant
91	qdrant_url: str = "http://localhost:6333"
92	qdrant_api_key: str \| None = None
93	collection_name: str = "documents"
94
95	# RAG settings
96	chunk_size: int = 1000
97	chunk_overlap: int = 150
98	top_k: int = 5
99	score_threshold: float = 0.7
100	max_context_tokens: int = 3000
101
102	# Redis (optional caching)
103	redis_url: str \| None = None
104
105	class Config:
106	env_file = ".env"
107
108	@lru_cache
109	def get_settings() -> Settings:
110	return Settings()
111	```
112
113	embeddings.py:
114	```python
115	from openai import OpenAI
116	from typing import List
117	import hashlib
118	import json
119
120	class EmbeddingService:
121	def __init__(self, settings):
122	self.client = OpenAI(api_key=settings.openai_api_key)
123	self.model = settings.embedding_model
124	self.cache = {} # Simple in-memory cache
125
126	def embed(self, text: str) -> List[float]:
127	"""Generate embedding for a single text."""
128	cache_key = hashlib.md5(text.encode()).hexdigest()
129	if cache_key in self.cache:
130	return self.cache[cache_key]
131
132	response = self.client.embeddings.create(
133	model=self.model,
134	input=text
135	)
136	embedding = response.data[0].embedding
137	self.cache[cache_key] = embedding
138	return embedding
139
140	def embed_batch(self, texts: List[str]) -> List[List[float]]:
141	"""Generate embeddings for multiple texts."""
142	response = self.client.embeddings.create(
143	model=self.model,
144	input=texts
145	)
146	return [item.embedding for item in response.data]
147	```
148
149	vectorstore.py:
150	```python
151	from qdrant_client import QdrantClient
152	from qdrant_client.models import (
153	VectorParams, Distance, PointStruct,
154	Filter, FieldCondition, MatchValue
155	)
156	from typing import List, Dict, Any
157	import uuid
158
159	class VectorStore:
160	def __init__(self, settings):
161	self.client = QdrantClient(
162	url=settings.qdrant_url,
163	api_key=settings.qdrant_api_key
164	)
165	self.collection_name = settings.collection_name
166	self.embedding_dim = 1536 # text-embedding-3-small
167
168	def create_collection(self):
169	"""Create collection if it doesn't exist."""
170	collections = self.client.get_collections().collections
171	exists = any(c.name == self.collection_name for c

People who liked this also viewed

6 picks

Type

azure-aiUse for Azure AI: Search, Speech, OpenAI, Document Intelligence. Helps with search, vector/hybrid search, speech-to-text, text-to-speech, transcription, OCR. WHEN: AI Search, query search, vector searSkillsJul 2026472k1.3k
just-scrapeSearch, scrape, crawl, extract structured data, and monitor web pages via the ScrapeGraph AI CLI. Use when the user asks to search the web, scrape a webpage, grab content from a URL, extract JSON fromSkillsJul 2026245k37
firecrawlCLI and Agent Skill for Firecrawl - Add scrape, search, and browsing capabilities to your AI agentsSkillsJul 202694k531
firecrawl-searchWeb search with full page content extraction. Use this skill whenever the user asks to search the web, find articles, research a topic, look something up, find recent news, discover sources, or says "SkillsJul 202676k531
firecrawl-agentAI-powered autonomous data extraction that navigates complex sites and returns structured JSON. Use this skill when the user wants structured data from websites, needs to extract pricing tiers, producSkillsJul 202674k531
firecrawl-mapDiscover and list all URLs on a website, with optional search filtering. Use this skill when the user wants to find a specific page on a large site, list all URLs, see the site structure, find where sSkillsJul 202674k531

rag-scaffold

33stars3forks

View on GitHub

How to install rag-scaffold?

floflo777/claude-rag-skills/rag-scaffold

$npx -y skills add floflo777/claude-rag-skills --skill rag-scaffold

Installs into the current project.

Summary

Generate production-ready RAG pipeline boilerplate code with best practices built-in.

Files · 1

View on GitHub

SKILL.md

1	# RAG Scaffold Skill
2
3	Generate production-ready RAG pipeline boilerplate code with best practices built-in.
4
5	## When to Use
6
7	Use `/rag-scaffold` when:
8	- Starting a new RAG project from scratch
9	- Need a reference implementation with best practices
10	- Want to quickly prototype a RAG system
11	- Learning RAG architecture patterns
12
13	## Scaffold Options
14
15	### Framework Choice
16	1. Python + LangChain - Most popular, extensive ecosystem
17	2. Python + LlamaIndex - Document-focused, great for complex pipelines
18	3. Python + Vanilla - No framework, full control
19	4. TypeScript + LangChain.js - For Node.js environments
20	5. Ailog API - Managed RAG-as-a-Service (simplest)
21
22	### Vector Store Choice
23	1. Qdrant - High performance, filtering, hybrid search
24	2. Pinecone - Managed, scalable, serverless option
25	3. ChromaDB - Simple, local-first, good for prototyping
26	4. Weaviate - GraphQL API, hybrid search
27	5. Milvus - High scale, GPU acceleration
28
29	### LLM Provider
30	1. OpenAI - GPT-4o, GPT-4o-mini
31	2. Anthropic - Claude 3.5 Sonnet, Claude 3 Opus
32	3. Mistral - Mistral Large, Mistral Small
33	4. Local - Ollama, vLLM
34
35	## How to Generate Scaffold
36
37	When the user invokes `/rag-scaffold`, ask:
38
39	1. What's your use case? (Customer support, documentation search, code assistant)
40	2. Preferred framework? (LangChain, LlamaIndex, Vanilla, Ailog API)
41	3. Vector store? (Qdrant, Pinecone, ChromaDB, etc.)
42	4. LLM provider? (OpenAI, Anthropic, Mistral)
43	5. Features needed?
44	- [ ] Hybrid search (dense + BM25)
45	- [ ] Reranking
46	- [ ] Conversation memory
47	- [ ] Streaming responses
48	- [ ] Source citations
49	- [ ] Multi-tenancy
50
51	## Scaffold Templates
52
53	### Template 1: Python + LangChain + Qdrant + OpenAI
54
55	Project Structure:
56	```
57	my-rag-project/
58	├── src/
59	│ ├── __init__.py
60	│ ├── config.py # Configuration management
61	│ ├── embeddings.py # Embedding service
62	│ ├── vectorstore.py # Vector store operations
63	│ ├── retriever.py # Retrieval logic
64	│ ├── generator.py # LLM generation
65	│ ├── rag_pipeline.py # Main RAG orchestration
66	│ └── chunker.py # Document chunking
67	├── scripts/
68	│ ├── index_documents.py # Indexing script
69	│ └── evaluate.py # Evaluation script
70	├── tests/
71	│ ├── test_retriever.py
72	│ └── test_generator.py
73	├── .env.example
74	├── requirements.txt
75	├── docker-compose.yml # Qdrant + Redis
76	└── README.md
77	```
78
79	config.py:
80	```python
81	from pydantic_settings import BaseSettings
82	from functools import lru_cache
83
84	class Settings(BaseSettings):
85	# OpenAI
86	openai_api_key: str
87	embedding_model: str = "text-embedding-3-small"
88	llm_model: str = "gpt-4o-mini"
89
90	# Qdrant
91	qdrant_url: str = "http://localhost:6333"
92	qdrant_api_key: str \| None = None
93	collection_name: str = "documents"
94
95	# RAG settings
96	chunk_size: int = 1000
97	chunk_overlap: int = 150
98	top_k: int = 5
99	score_threshold: float = 0.7
100	max_context_tokens: int = 3000
101
102	# Redis (optional caching)
103	redis_url: str \| None = None
104
105	class Config:
106	env_file = ".env"
107
108	@lru_cache
109	def get_settings() -> Settings:
110	return Settings()
111	```
112
113	embeddings.py:
114	```python
115	from openai import OpenAI
116	from typing import List
117	import hashlib
118	import json
119
120	class EmbeddingService:
121	def __init__(self, settings):
122	self.client = OpenAI(api_key=settings.openai_api_key)
123	self.model = settings.embedding_model
124	self.cache = {} # Simple in-memory cache
125
126	def embed(self, text: str) -> List[float]:
127	"""Generate embedding for a single text."""
128	cache_key = hashlib.md5(text.encode()).hexdigest()
129	if cache_key in self.cache:
130	return self.cache[cache_key]
131
132	response = self.client.embeddings.create(
133	model=self.model,
134	input=text
135	)
136	embedding = response.data[0].embedding
137	self.cache[cache_key] = embedding
138	return embedding
139
140	def embed_batch(self, texts: List[str]) -> List[List[float]]:
141	"""Generate embeddings for multiple texts."""
142	response = self.client.embeddings.create(
143	model=self.model,
144	input=texts
145	)
146	return [item.embedding for item in response.data]
147	```
148
149	vectorstore.py:
150	```python
151	from qdrant_client import QdrantClient
152	from qdrant_client.models import (
153	VectorParams, Distance, PointStruct,
154	Filter, FieldCondition, MatchValue
155	)
156	from typing import List, Dict, Any
157	import uuid
158
159	class VectorStore:
160	def __init__(self, settings):
161	self.client = QdrantClient(
162	url=settings.qdrant_url,
163	api_key=settings.qdrant_api_key
164	)
165	self.collection_name = settings.collection_name
166	self.embedding_dim = 1536 # text-embedding-3-small
167
168	def create_collection(self):
169	"""Create collection if it doesn't exist."""
170	collections = self.client.get_collections().collections
171	exists = any(c.name == self.collection_name for c

People who liked this also viewed

6 picks

Type