DE / app.py
rajrakeshdr's picture
Update app.py
c7eac83 verified
raw
history blame
6.89 kB
from fastapi import FastAPI, HTTPException
from pydantic import BaseModel
import requests
import os
from datetime import datetime, timedelta
from groq import Groq
from dotenv import load_dotenv
import logging
# Configure logging
logging.basicConfig(level=logging.DEBUG)
logger = logging.getLogger(__name__)
# Load environment variables
load_dotenv()
# Configuration (Use environment variables instead of hardcoding)
GITHUB_TOKEN = "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"
GROQ_API_KEY = "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"
REPOSITORIES = [
"falcosecurity/rules",
"SigmaHQ/sigma",
"reversinglabs/reversinglabs-yara-rules",
"elastic/detection-rules",
"sublime-security/sublime-rules",
"Yamato-Security/hayabusa-rules",
"anvilogic-forge/armory",
"chainguard-dev/osquery-defense-kit",
"splunk/security_content",
"Neo23x0/signature-base",
"SlimKQL/Hunting-Queries-Detection-Rules"
]
DAYS_BACK = 1
# GitHub API base URL
GITHUB_API_URL = "https://api.github.com"
# Groq client setup
groq_client = Groq(api_key=GROQ_API_KEY)
# FastAPI app
app = FastAPI(docs_url=None, redoc_url=None)
class RepositoryDetails(BaseModel):
repo_name: str
repo_url: str
changes: str
description: str
context: str
def fetch_repository_changes(repo: str, days_back: int) -> list[str]:
"""
Fetch recent commits and pull requests for a repository.
"""
try:
logger.debug(f"Fetching changes for repository: {repo}")
since_date = (datetime.now() - timedelta(days=days_back)).isoformat()
headers = {
"Authorization": f"token {GITHUB_TOKEN}",
"Accept": "application/vnd.github.v3+json"
}
# Fetch commits
commits_url = f"{GITHUB_API_URL}/repos/{repo}/commits"
commits_params = {"since": since_date}
logger.debug(f"Fetching commits from: {commits_url} with params: {commits_params}")
commits_response = requests.get(commits_url, headers=headers, params=commits_params)
commits_response.raise_for_status()
commits = commits_response.json()
logger.debug(f"Found {len(commits)} commits for {repo}")
# Fetch pull requests
prs_url = f"{GITHUB_API_URL}/repos/{repo}/pulls"
prs_params = {"state": "all", "sort": "updated", "direction": "desc"}
logger.debug(f"Fetching pull requests from: {prs_url} with params: {prs_params}")
prs_response = requests.get(prs_url, headers=headers, params=prs_params)
prs_response.raise_for_status()
prs = prs_response.json()
logger.debug(f"Found {len(prs)} pull requests for {repo}")
# Extract changes
changes = []
for commit in commits:
changes.append(f"Commit: {commit['commit']['message']}")
for pr in prs:
updated_at = datetime.strptime(pr["updated_at"], "%Y-%m-%dT%H:%M:%SZ")
if updated_at >= datetime.now() - timedelta(days=days_back):
changes.append(f"PR: {pr['title']} - {pr['body'] or 'No description'}")
logger.debug(f"Total changes for {repo}: {len(changes)}")
return changes
except requests.exceptions.RequestException as e:
logger.error(f"Error fetching changes for {repo}: {e}")
raise HTTPException(status_code=500, detail=f"Error fetching changes for {repo}: {e}")
def summarize_changes_with_deepseek(repo: str, changes: list[str]) -> dict:
"""
Use Groq's DeepSeek model to summarize changes and provide insights.
"""
try:
logger.debug(f"Summarizing changes for repository: {repo}")
prompt = f"""
Analyze the following changes made to detection rules in the GitHub repository {repo}:
{', '.join(changes)}
Provide a detailed response with two sections:
- Description: Summarize what changes were made.
- Context: Explain why these changes might be required.
"""
logger.debug(f"Sending prompt to DeepSeek: {prompt[:100]}...") # Truncate for brevity in logs
response = groq_client.chat.completions.create(
model="deepseek-chat",
messages=[{"role": "user", "content": prompt}],
max_tokens=500,
temperature=0.7
)
summary = response.choices[0].message.content
logger.debug(f"Received summary from DeepSeek: {summary[:100]}...")
# Extract description and context with fallback
description = "Description not found."
context = "Context not found."
if "Description:" in summary and "Context:" in summary:
description = summary.split("Description:")[1].split("Context:")[0].strip()
context = summary.split("Context:")[1].strip()
else:
description = summary # Fallback to full summary if sections aren't clear
return {"description": description, "context": context}
except Exception as e:
logger.error(f"Error summarizing changes for {repo}: {e}")
raise HTTPException(status_code=500, detail=f"Error summarizing changes for {repo}: {e}")
@app.get("/monitor", response_model=list[RepositoryDetails])
async def monitor_repositories():
"""
Single API endpoint to fetch and summarize changes for all repositories.
"""
try:
logger.debug("Starting to monitor repositories")
results = []
for repo in REPOSITORIES:
logger.debug(f"Processing repository: {repo}")
changes = fetch_repository_changes(repo, DAYS_BACK)
if changes:
logger.debug(f"Summarizing changes for {repo}")
summary = summarize_changes_with_deepseek(repo, changes)
results.append(RepositoryDetails(
repo_name=f"{repo} (+{len(changes)})",
repo_url=f"https://github.com/{repo}",
changes="\n".join(changes),
description=summary["description"],
context=summary["context"]
))
else:
logger.debug(f"No changes detected for {repo}")
results.append(RepositoryDetails(
repo_name=f"{repo} (No changes)",
repo_url=f"https://github.com/{repo}",
changes="No changes detected in the last 7 days.",
description="No changes detected.",
context="No context available."
))
logger.debug("Finished monitoring repositories")
return results
except Exception as e:
logger.error(f"Error in monitor_repositories: {e}")
raise HTTPException(status_code=500, detail=str(e))
if __name__ == "__main__":
import uvicorn
uvicorn.run(app, host="0.0.0.0", port=8000)