How watchtowrlabs/watchtowr-vs-ivanti-epmm-cve-2025-4427-cve-2025-4428 works
Project type and purpose
• This is a small command-line detection tool written in Python.
• Purpose: send crafted HTTP requests to an Ivanti EPMM server and determine if it is vulnerable to the CVE-2025-4427 + CVE-2025-4428 pre-auth RCE chain. It generates an “artifact” (a proof-of-concept request/response sequence) and prints vulnerability status.
File structure
• Top level contains:
• README.md – project description and usage
• watchTowr-vs-Ivanti-EPMM-CVE-2025-4427-CVE-2025-4428.py – the main script
• No subdirectories, no external configs, single-module script.
README for additional insights
• Confirms the script runs via “python3 <script> –H <target>.”
• Describes the detection steps (e.g., sending id command).
• No external microservices, pure client→server interaction.
Main components
a. CLI Frontend
– Parses command-line argument for target URL
– Options parser component (e.g., argparse)
b. HTTP Client
– Builds and sends HTTP requests to /vulnerable endpoint
– Implements any necessary headers or payload encoding
– Uses Python requests library or http.client
c. Vulnerability Detection Module
– Encapsulates the logic of chaining CVE-2025-4427 + CVE-2025-4428 exploits
– Crafts proof-of-concept payloads (e.g., injecting shell commands)
– Parses and inspects HTTP responses for indicators of RCE (e.g., the output of id)
d. Output / Reporter
– Interprets the detection module’s result
– Prints “VULNERABLE” or “NOT VULNERABLE” to the console
– Optionally writes an artifact file or log
e. External Target Server
– Ivanti EPMM service under test
– Shows the network boundary between your script and the target
Relationships and interactions
• The CLI Frontend feeds the target URL into the HTTP Client.
• The HTTP Client sends requests over the network to the Ivanti EPMM server.
• Responses come back to the Vulnerability Detection Module.
• Detection Module analyzes response, returns a boolean or status string.
• Output / Reporter prints the final verdict to standard output.
Architectural patterns and design principles
• Single-process, linear flow (no concurrency required).
• Modular functions: separation of concerns (parsing, HTTP, detection, reporting).
• External dependency: Python standard library plus any HTTP library.
Detailed capture of essential elements
• Python runtime as an overall container.
• Optional subprocess or shell invocation if the script uses subprocess for local analysis.
• Detection steps: “Exploit step 1 (CVE-4427) → session established → Exploit step 2 (CVE-4428).”
• That the HTTP Client and Detector collaborate to form the RCE chain.
Final advice
• Use precise labels: component names and data formats (JSON, headers, shell output).