Hugging Face's logo

Spaces:
Prakyath01
/
kubernetes-rag-assistant
Sleeping

App Files Community
kubernetes-rag-assistant
File size: 4,952 Bytes 
4d068e8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
import os
import requests
import json
from bs4 import BeautifulSoup
from textwrap import shorten

import gradio as gr

from langchain_core.documents import Document
from langchain_text_splitters import RecursiveCharacterTextSplitter
from langchain_community.vectorstores import Chroma
from langchain_community.embeddings import HuggingFaceEmbeddings

# -----------------------
# 1. SCRAPE K8S DOCS
# -----------------------
urls = {
    "pods": "https://kubernetes.io/docs/concepts/workloads/pods/",
    "deployments": "https://kubernetes.io/docs/concepts/workloads/controllers/deployment/",
    "services": "https://kubernetes.io/docs/concepts/services-networking/service/",
    "namespaces": "https://kubernetes.io/docs/concepts/overview/working-with-objects/namespaces/",
    "nodes": "https://kubernetes.io/docs/concepts/architecture/nodes/",
    "statefulsets": "https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/",
    "rbac": "https://kubernetes.io/docs/reference/access-authn-authz/rbac/",
    "persistent-volumes": "https://kubernetes.io/docs/concepts/storage/persistent-volumes/",
    "ingress": "https://kubernetes.io/docs/concepts/services-networking/ingress/",
    "autoscaling": "https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/"
}

def scrape_docs():
    docs = []
    for name, url in urls.items():
        try:
            r = requests.get(url, timeout=20)
            soup = BeautifulSoup(r.text, "html.parser")
            content = soup.find("div", class_="td-content")
            if not content:
                continue
            text = content.get_text(separator="\n").strip()
            docs.append(Document(page_content=text, metadata={"doc_id": name, "url": url}))
        except Exception:
            continue
    return docs

docs = scrape_docs()

# -----------------------
# 2. CHUNK + EMBED + VECTOR DB
# -----------------------
splitter = RecursiveCharacterTextSplitter(chunk_size=800, chunk_overlap=100)
chunks = splitter.split_documents(docs)

embedding = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
vectordb = Chroma.from_documents(chunks, embedding)
retriever = vectordb.as_retriever(
    search_type="similarity_score_threshold",
    search_kwargs={"k": 5, "score_threshold": 0.4}
)

# -----------------------
# 3. RAG HELPERS
# -----------------------
def build_context_with_citations(query: str):
    retrieved_docs = retriever.invoke(query)
    context = ""
    mapping = []

    for i, d in enumerate(retrieved_docs, start=1):
        label = f"[{i}]"
        context += f"{label} {d.page_content[:1000]}\n\nSource: {d.metadata['url']}\n\n"
        mapping.append({
            "label": label,
            "url": d.metadata["url"],
            "doc": d.metadata["doc_id"],
            "preview": shorten(d.page_content, width=200)
        })
    return context, mapping

def build_prompt(query, context):
    return f"""
You are a Kubernetes expert.
Use ONLY the context below.
Add citations like [1][2] after each fact.
If not found, say: 'Not in docs'.

QUESTION:
{query}

CONTEXT:
{context}
""".strip()

# -----------------------
# 4. OPENROUTER LLM
# -----------------------
import requests as req

OPENROUTER_API_KEY = os.environ.get("OPENROUTER_API_KEY", "")

def call_llm(prompt: str) -> str:
    if not OPENROUTER_API_KEY:
        return "OpenRouter API key is not set. Please configure OPENROUTER_API_KEY in the Space settings."

    url = "https://openrouter.ai/api/v1/chat/completions"
    headers = {
        "Authorization": f"Bearer {OPENROUTER_API_KEY}",
        "Content-Type": "application/json"
    }
    data = {
        "model": "meta-llama/llama-3.1-8b-instruct",
        "messages": [
            {"role": "system", "content": "You are a Kubernetes expert. Only use provided context."},
            {"role": "user", "content": prompt}
        ],
        "temperature": 0.0
    }
    response = req.post(url, headers=headers, data=json.dumps(data))
    out = response.json()
    return out.get("choices", [{"message": {"content": "No response"}}])[0]["message"]["content"]

def answer_question(query: str):
    context, sources = build_context_with_citations(query)
    prompt = build_prompt(query, context)
    answer = call_llm(prompt)
    return answer, sources

# -----------------------
# 5. GRADIO CHAT APP
# -----------------------
def chat_fn(message, history):
    answer, sources = answer_question(message)
    src_lines = [f"{s['label']} – {s['url']}" for s in sources]
    sources_text = "\n".join(src_lines) if src_lines else "No sources found."
    full_answer = f"{answer}\n\n---\nSources:\n{sources_text}"
    return full_answer

demo = gr.ChatInterface(
    fn=chat_fn,
    title="Kubernetes RAG Assistant",
    description="Ask Kubernetes questions. Answers are grounded in official docs and include citations."
)

def main():
    return demo

if __name__ == "__main__":
    demo.launch()