Hugging Face's logo

Spaces:
starlord3307
/
docker
Sleeping

App Files Community
docker
File size: 4,727 Bytes
c7b1cc2
3f8c649
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1eeb956
 
 
3f8c649
1eeb956
 
 
3f8c649
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
 
import os
from fastapi import FastAPI, Request
from huggingface_hub import InferenceClient
import json

app = FastAPI()

HF_API_KEY = os.getenv("HF_API_KEY")  # 🔥 No Hardcoded Keys!

# ✅ Set up Hugging Face Inference Client
client = InferenceClient(
    provider="hf-inference",
    api_key=HF_API_KEY  # Use the key from environment
)

# ✅ Tag descriptions (helps model understand each tag's meaning)
TAGS_DESCRIPTION = """
IPWhitelisting: Adding IP addresses into Okta Security Networks fields if blocked.
AppLocker: Used when adding hashes to allow execution of applications (AppLocker or CrowdStrike).
ADSecurityGroup: Removing a user from an AD Security group to stop or allow access.
AttachmentRelease: Releasing an attachment if password protected.
EmailWhitelisting: Allowing emails through ProofPoint or Checkpoint for delivery.
AppAssignment: Adding a user to an Application.
chatops:mfa-bypass: Automates MFA AD group changes or Requests from the bypass MFA.
VM: Removing hosts from Nessus or decommissioning hosts from tracking in nessus and tenable or All requests for tenable and nessus.
PhishingReport: When a user submits a phishing report or Email investigation or Email Analysis.
GenericInformation: Generic tickets assigned to SecOps.
PasswordReset: When a ticket needs Okta or AD password reset.
KeeperAccounts: Issues with Keeper Password Manager accounts.
MemberIssues: Investigating possible security issues with Member accounts.
ADPassword: AD password issues, including resets.
OktaMFAResets: Okta MFA resets performed by Security or Requests for MFA reset.
Zscaler: Issues related to Zscaler security or Domain/URL allow requests or Domain/URL block requests.
chatops:cs-usb: Automates individual USB whitelisting.
USBDeviceControl: Requests for workstation USB whitelisting.
Imperva: Issues related to Imperva and CDN security.
"""


# ✅ Function to summarize & assign a single tag
def analyze_ticket(title, body):
    messages = [
        {"role": "system",
         "content": "You are an AI that summarizes security tickets and assigns a single most relevant tag based on "
                    "the title and body."},
        {"role": "user", "content": f"Title: {title}\nBody: {body}\n\n"
                                    "### TASK 1: Summarization\nSummarize the ticket body in 1-2 sentences.\n\n"
                                    "### TASK 2: Tagging\nChoose the **one best tag** from this list that matches the "
                                    "ticket:\n "
                                    f"{TAGS_DESCRIPTION}\n\n"
                                    "Return output in JSON format with keys: 'summary' (ticket summary) and 'tag' ("
                                    "best matching tag)."}
    ]

    completion = client.chat.completions.create(
        model="mistralai/Mistral-7B-Instruct-v0.3",
        messages=messages,
        max_tokens=200
    )

    # ✅ Parse JSON response
    try:
        response = json.loads(completion.choices[0].message.content.strip())
    except json.JSONDecodeError:
        response = {"summary": "Could not generate summary", "tag": "Unknown"}

    return response


# ✅ API Endpoint to accept input from Tines
@app.post("/process-ticket/")
async def process_ticket(request: Request):
    data = await request.json()

    # ✅ Extract from "inputs" instead of top-level JSON
    inputs = data.get("inputs", {})
    
    # ✅ Extract input fields
    ticket_id = inputs.get("id", "No ID")
    title = inputs.get("title", "No Title")
    body = inputs.get("body", "No Body")

    # ✅ Get AI-generated summary & tag
    result = analyze_ticket(title, body)

    # ✅ Return response including ID
    return {
        "id": ticket_id,  # ✅ Keep original ID
        "summary": result["summary"],
        "tag": result["tag"]
    }

# def process_input_file(filename):
#     try:
#         with open(filename, "r", encoding="utf-8") as file:
#             data = json.load(file)  # ✅ Load JSON file
#
#         title = data.get("title", "No Title")
#         body = data.get("body", "No Body")
#
#         # ✅ Get AI-generated summary & tag
#         result = analyze_ticket(title, body)
#
#         return result  # ✅ Returns {'summary': "...", 'tag': "..."}
#
#     except FileNotFoundError:
#         print(f"❌ Error: The file '{filename}' was not found.")
#         return {"error": "File not found"}
#
#     except json.JSONDecodeError:
#         print(f"❌ Error: The file '{filename}' contains invalid JSON.")
#         return {"error": "Invalid JSON format"}
#
#
# # ✅ Load input from `input.json`
# output = process_input_file("input.json")
#
# # ✅ Print the output
# print(json.dumps(output, indent=2))