Exploit catalog results

Showing 2 PoCs on this page

GitHub

jimmexploit/CVE-2024-37054-PoC

Repository PoCStars: 6Created 2026-05-18
ExploitCVE-2024-370544 files

8.8 KiB

Analysisdeepseek-v4-pro:cloud ·

Technical assessment

Full exploit for CVE-2024-37054 targeting MLflow. The code registers an account, trains a model, uploads a malicious pickle payload via the MLflow artifact API, and triggers deserialization to achieve remote code execution via a reverse shell.

Backdoor review

No backdoor observed in reviewed code

The supplied evidence is a proof-of-concept exploit for CVE-2024-37054, a deserialization vulnerability in MLflow. The code builds a malicious pickle payload that executes a reverse shell when loaded by the vulnerable MLflow server. All behavior is consistent with the stated exploit objective: triggering arbitrary code execution on the target MLflow instance. No concealed backdoor, unrelated payload, or operator-directed harm beyond the advertised exploit was observed.

ClassificationExploit
Model confidence98%
AuthenticationRequired
LanguagesPython
Target softwareMLflow
Attack typesRemote Code ExecutionDeserialization of Untrusted Data
Evidence & reasoningClassification basis · observed behavior · safety review
Technical evidence

Classification basis and observed behavior

Classification basis

The artifact contains a complete, multi-step exploit that actively exercises the CVE-2024-37054 vulnerability to achieve remote code execution. It includes account registration, payload generation, artifact upload, and trigger mechanisms, going beyond mere detection or scanning.

shell.py:1-185README.MD:3-4

Requirements

  • Network access to a target MLflow instance and associated web application.shell.py:26-27
  • Valid credentials for the MLflow API (default admin:password) or ability to register an account on the target application.shell.py:31-32shell.py:40-76
  • A listener (e.g., netcat) to receive the reverse shell connection.shell.py:181

Observed behavior

  • Registers a new user account on the target web application and retrieves a session cookie.shell.py:40-76
  • Uploads a CSV file to trigger model training and retrieves the MLflow run_id.shell.py:78-116
  • Constructs a malicious pickle payload that executes a bash reverse shell command upon deserialization.shell.py:118-128
  • Uploads the malicious pickle to the MLflow artifact store, overwriting the model's python_model.pkl.shell.py:130-146
  • Triggers the /predict endpoint on the target application, causing the malicious model to be loaded and the reverse shell to execute.shell.py:148-159
Safety-review evidence

Behaviors behind the backdoor verdict

Observables

Reverse Shell Payload
Payload withheldThe exploit constructs a reverse shell command using attacker-supplied LHOST and LPORT, which is the expected behavior for a CVE-2024-37054 PoC.shell.py:120malicious_payload.py:13
Pickle Deserialization Exploit
Payload withheldThe malicious pickle payload is designed to execute os.system with the reverse shell command upon deserialization, directly implementing the CVE-2024-37054 vulnerability.shell.py:122-126malicious_payload.py:15-17
Target Interaction
Payload withheldThe exploit interacts with the target application and MLflow server to upload the malicious model and trigger deserialization, all within the scope of the described exploit chain.shell.py:50-55shell.py:64-69shell.py:84-88shell.py:136-141shell.py:154-158
Review boundaries

What the analysis did not establish

  • Evidence is limited to the source code; no runtime behavior or network traffic was observed.
  • The exploit targets a specific lab environment (HTB SmartHire) and may require adaptation for other targets.
  • Only the readable text files were reviewed; no binary files were present. The analysis assumes the provided source code is the complete artifact and that no external dependencies introduce hidden behavior.
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

jimmexploit/CVE-2026-42167-PoC

Repository PoCStars: 0Created 2026-05-02
ExploitCVE-2026-421674 files

7.4 KiB

Analysisdeepseek-v4-pro:cloud ·

Technical assessment

A Go-based exploit for CVE-2026-42167 that performs pre-authentication SQL injection via the FTP USER command against ProFTPD's mod_sql. It offers two modes: 'payload withheld' to inject a new user into the database, and 'payload withheld' to execute a reverse shell using PostgreSQL's COPY TO PROGRAM.

Backdoor review

No backdoor observed in reviewed code

The supplied evidence consists of a Go-based proof-of-concept exploit for CVE-2026-42167. The code implements two attack modes: a 'payload withheld' mode that injects a new FTP user via SQL injection, and an 'payload withheld' mode that attempts to execute a reverse shell via a COPY TO PROGRAM SQL command. Both modes target the stated vulnerability in ProFTPD and perform no actions beyond demonstrating the exploit against a user-supplied target. No concealed executable behavior, persistence mechanisms, credential exfiltration, or unrelated payloads were observed.

ClassificationExploit
Model confidence95%
AuthenticationNot required
LanguagesGo
Target softwareProFTPD
Attack typesSQL InjectionRemote Code Execution
Evidence & reasoningClassification basis · observed behavior · safety review
Technical evidence

Classification basis and observed behavior

Classification basis

The code actively exploits the vulnerability by sending malicious SQL payloads via the FTP USER command to achieve unauthorized actions (user creation and remote command execution). It is not merely detecting the vulnerability but exercising it to gain access or execute code.

main.go:82-83main.go:150README.md:7

Requirements

  • Target ProFTPD server must be running a vulnerable version (1.3.9 and below) with mod_sql enabled and configured to log USER requests with an expansion like %U.README.md:7README.md:10
  • The SQL backend must allow commands such as COPY TO PROGRAM (for payload withheld mode) or INSERT (for payload withheld mode).main.go:82-83main.go:150
  • For payload withheld mode, a listener must be set up to catch the reverse shell.main.go:149main.go:188-189

Observed behavior

  • Connects to the target FTP server and sends a crafted USER command containing a SQL injection payload.main.go:86-90main.go:158-162
  • In payload withheld mode, the payload is an INSERT statement to create a new FTP user with root privileges.main.go:82-83
  • In payload withheld mode, the payload uses COPY TO PROGRAM to execute a bash reverse shell command.main.go:149-150
  • After injection, the payload withheld mode attempts to authenticate with the newly created user to confirm success.main.go:95-101
Safety-review evidence

Behaviors behind the backdoor verdict

Observables

Exploit Mode
Payload withheldThe 'payload withheld' mode injects a new FTP user into the target database via SQL injection, then attempts to log in with the injected credentials. This is a direct demonstration of the CVE-2026-42167 vulnerability.main.go:54-133
Exploit Mode
Payload withheldThe 'payload withheld' mode sends a SQL injection payload containing a COPY TO PROGRAM command to execute a reverse shell on the target. This is a direct demonstration of the CVE-2026-42167 vulnerability.main.go:135-168
Network Connection
Payload withheldThe exploit connects only to the target FTP server specified by the user via command-line flags. No other network connections are initiated.main.go:181-182main.go:13-15
Reverse Shell
Payload withheldIn 'payload withheld' mode, the exploit constructs a reverse shell command that connects back to a listener address and port provided by the user. This is a standard technique for demonstrating remote code execution and is not a hidden payload withheld.main.go:149-150main.go:188-189
Review boundaries

What the analysis did not establish

  • The file go.sum is omitted from the evidence, so dependency integrity cannot be verified.
  • The evidence does not include any network captures or execution output confirming the exploit was successful in a live environment.
  • The file go.sum was not included as readable text; it is a dependency checksum file and unlikely to contain executable behavior.
  • The analysis is limited to the supplied text evidence; no dynamic analysis or execution was performed.
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.