Skip to content
Submit incident
AII-1001-85391
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

DoS vulnerability in Applio restart.py via arbitrary process termination

Applio version 3.2.7 is vulnerable to denial of service through a flaw in restart.py where user input for model_name is used unsanitized to construct file paths and access process IDs for termination. An attacker can exploit path traversal and arbitrary file writes to kill critical processes and crash the application.

11 December 2024 Β· Applio Β· Security and cyberattacks
11 December 2024
AII-1001-853913 sources ↓
Quick facts
Category
Security and cyberattacks
Source
github.com
Date
11 December 2024
Drafted by
Raidu Incident Detection

What happened

The stop_train function in restart.py takes user input for model_name without sanitization and uses it to construct a file path to access config.json.

An attacker can use path traversal (e.g., supplying "../../" in model_name) to read process IDs from config.json files in arbitrary locations, or combine this with arbitrary file writes to plant a config.json containing hundreds of attacker-controlled process IDs.

When the vulnerable endpoint is accessed, the function reads these process IDs and terminates them using os.kill, potentially killing the Applio process itself or other critical processes on the system.

Reported impact

The vulnerability enables denial of service by allowing an attacker to terminate critical processes and crash the Applio application.

Index analysis

The root cause is the use of unsanitized user input (model_name) in os.path.join to construct file paths, enabling both path traversal and arbitrary process termination.

The broad exception handling (except: pass) masks errors and allows the attack to proceed silently.

Governance controls

GitHub Security Lab reported the vulnerability to the Applio maintainers on 2024-12-11 and conducted a coordinated disclosure process over three months, with multiple follow-ups.

Classification

Category
Security and cyberattacks
Industry
General or personal use
Harm status
Potential harm
Intent
Unintentional
Timing
Post-deployment
Failure modes
Unsafe autonomous behaviour
Consequences
System suspension or termination
AI systems
Applio

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

github.com
https://github.com/IAHispano/Applio/blob/29b4a00e4be209f9aac51cd9ccffcc632dfb2973/tabs/settings/sections/restart.py#L9
securitylab.github.com
https://securitylab.github.com/advisories/GHSL-2024-354_Applio/
cve.org
NVD entry