Exploit catalog results

Showing 2 PoCs on this page

GitHub

Dhiaelhak-Rached/CVE-2026-34753

Repository PoCStars: 0Created 2026-05-03
ExploitCVE-2026-347538 files

5.4 KiB

Analysisdeepseek-v4-pro:cloud ·

Technical assessment

The artifact is a lab environment that demonstrates exploitation of CVE-2026-34753, an SSRF vulnerability in vLLM. It includes a vulnerable mock server (victim/app.py) that replicates the vulnerable download_bytes_from_url function, an internal mock metadata service (victim/metadata.py), and an attacker script (attacker/attack.sh) that sends a crafted JSON payload to trigger the SSRF and fetch data from the internal service.

Backdoor review

No backdoor observed in reviewed code

The repository contains a proof-of-concept (PoC) for CVE-2026-34753, an SSRF vulnerability in vLLM. The attacker script (attack.sh) sends a crafted request to a local victim service, and the victim code (app.py) demonstrates the vulnerable function. The metadata.py file simulates an internal metadata service. No backdoor, trojan, or deceptive behavior targeting the PoC operator was observed; all code serves the stated educational/demonstration purpose.

ClassificationExploit
Model confidence95%
AuthenticationNot required
LanguagesPythonShell
Target softwarevLLM
Attack typesServer-Side Request Forgery (SSRF)
Evidence & reasoningClassification basis · observed behavior · safety review
Technical evidence

Classification basis and observed behavior

Classification basis

The artifact contains an attacker script (attack.sh) that sends a crafted request to a vulnerable endpoint, and a victim server (app.py) that executes the vulnerable code path to fetch attacker-controlled URLs. This constitutes an exploit because it exercises the vulnerability to cause the server to make requests to internal services, which is the intended malicious action.

attacker/attack.sh:14-17victim/app.py:27-36victim/app.py:57

Requirements

  • The attacker must be able to send HTTP requests to the vulnerable vLLM batch endpoint (/v1/batch).attacker/attack.sh:15
  • The attacker must control the 'file_url' field in the batch input JSON.victim/app.py:52

Observed behavior

  • The attacker script sends a POST request to the victim's /v1/batch endpoint with a JSON payload containing a file_url pointing to an internal service (e.g., http://metadata:8080/latest/meta-data/iam/security-credentials/).attacker/attack.sh:14-17
  • The vulnerable server makes an HTTP GET request to the attacker-supplied URL without any validation, fetches the response, and returns a preview of the fetched data in its JSON response.victim/app.py:27-36victim/app.py:57-69
  • The internal metadata service simulates an AWS EC2 metadata endpoint and returns mock credentials when queried.victim/metadata.py:6-14
Safety-review evidence

Behaviors behind the backdoor verdict

Observables

Url
Payload withheldTarget URL for the SSRF demonstration, hardcoded in attack.sh and app.py.attacker/attack.sh:3victim/app.py:83
Url
Payload withheldSimulated internal metadata service endpoint, used to demonstrate SSRF impact.victim/metadata.py:21
Credential
Payload withheldExample AWS credentials returned by the simulated metadata service; clearly fake and for demonstration only.victim/metadata.py:10-11
Review boundaries

What the analysis did not establish

  • The evidence does not include the actual exploit.jsonl payload file referenced by attack.sh, so the exact content of the SSRF request is not visible.
  • The victim server is a mock implementation, not the actual vLLM codebase, so the exploit's behavior against a real vLLM instance is not directly observed.
  • The evidence is a lab setup and does not demonstrate exploitation against a live, production system.
  • Four additional files (total 8) were omitted from the text evidence; their metadata indicates they are non-text or not selected, but their content was not reviewed.
  • Binary content, if any, was flagged as metadata-only and not analyzed.
Model interpretation

This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.

GitHub

Dhiaelhak-Rached/CVE-2026-39987-lab-or-marimo-cve-lab

Repository PoCStars: 0Created 2026-04-26
ExploitCVE-2026-399874 files

17.9 KiB

Analysisdeepseek-v4-pro:cloud ·

Technical assessment

The artifact contains a Python script (exploit.py) that connects to a marimo terminal WebSocket endpoint without authentication and executes arbitrary system commands or provides an interactive shell, directly exercising CVE-2026-39987.

Backdoor review

No backdoor observed in reviewed code

The supplied evidence consists of a README.md lab guide and an exploit.py script for CVE-2026-39987. The README describes the vulnerability and provides educational instructions for reproducing the authentication bypass in a Docker lab. The exploit.py script implements a proof-of-concept that connects to the vulnerable /terminal/ws WebSocket endpoint without authentication and executes arbitrary commands or provides an interactive shell. Both files are consistent with a legitimate educational PoC for the described CVE; no concealed backdoor, deceptive payload, or operator-directed harm is observed.

ClassificationExploit
Model confidence100%
AuthenticationNot required
LanguagesPython
Target softwaremarimo
Attack typesAuthentication BypassRemote Code Execution
Evidence & reasoningClassification basis · observed behavior · safety review
Technical evidence

Classification basis and observed behavior

Classification basis

The file exploit.py is a functional script that actively exploits the vulnerability by establishing an unauthenticated WebSocket connection to /terminal/ws and executing arbitrary commands, which is the definition of exploit code.

exploit.py:1-8exploit.py:16-78

Requirements

  • A vulnerable marimo instance (<= 0.20.4) running in edit mode with the /terminal/ws endpoint exposed.README.md:30
  • Python 3 with the websocket-client library installed.README.md:80

Observed behavior

  • Connects to a WebSocket URL without providing any authentication token.exploit.py:21-22
  • Sends a system command over the WebSocket and prints the output.exploit.py:44-47
  • Provides an interactive shell by reading user input and sending it as commands over the WebSocket.exploit.py:119-129
Safety-review evidence

Behaviors behind the backdoor verdict

Observables

Websocket Url
Payload withheldTarget WebSocket endpoint used by the PoC exploit to demonstrate the authentication bypass.README.md:81exploit.py:178
Command Execution
Payload withheldExample command executed by the PoC to demonstrate arbitrary command execution on the target.README.md:81
Review boundaries

What the analysis did not establish

  • The evidence includes only two of four repository files; Dockerfile.target and docker-compose.yml are omitted, but their absence does not affect classification of the included exploit.py.
  • The artifact code was not executed; classification is based solely on static analysis of the provided source text.
  • Two additional files (Dockerfile.target and docker-compose.yml) are present in the repository but were not included in the text evidence; their content could not be reviewed.
  • Binary content was not analyzed per the evidence envelope policy; no binary files were flagged.
Model interpretation

This review is limited to the supplied PoC code and context. It does not assert that the code works or is safe to execute.