CWE-1295

Backpropagate is a Python library for fine-tuning large language models on a single GPU. In versions 1.1.0 and 1.1.1, the optional Reflex web UI exposes a training control plane without authentication: dataset upload, model load, training start/stop, multi-run orchestration, GGUF export, and HuggingFace Hub push. The CLI accepts two operator-facing flags intended as security controls: --auth user:pass — documented as "require HTTP Basic authentication on every request to the UI." and--share — documented as "expose the UI on a public address; requires --auth." When --auth user:pass is passed, the CLI prints Auth: enabled (user: <username>) to confirm to the operator that authentication is active, then exports BACKPROPAGATE_UI_AUTH=user:pass to the subprocess that launches the Reflex backend. The Reflex backend (backpropagate/ui_app/**) never reads BACKPROPAGATE_UI_AUTH. No authentication middleware is registered. No request-level guard runs. No WebSocket upgrade guard runs. Any client that reaches the bound port — local or remote, depending on whether --share is used — has full UI access. An inline comment at backpropagate/cli.py:1217-1218 in the v1.1.0 source documents the gap: "For Phase 1 the variable is exported but Reflex doesn't read it yet." This comment was internal-facing; the user-facing documentation (README, CHANGELOG, SHIP_GATE) advertised the contract as enforced. An attacker who reaches the bound port can read uploaded datasets, trigger arbitrary training runs against any local base models as well as read their paths, trigger HuggingFace Hub pushes and cause disk-fill DoS. This issue has been fixed in version 1.2.0. If developers cannot immediately upgrade to 1.2.0 run backprop ui with no flags so it binds to localhost, use SSH port-forwarding (ssh -L 7860:localhost:7860 <training-host>) instead of --share for remote access, and audit any host previously launched with --share, re-issuing any HF tokens used during those sessions.

ai-client-html is an Aimeos e-commerce HTML client component. Debug information revealed sensitive information from environment variables in error log. This issue has been patched in versions 2024.04.7, 2023.10.15, 2022.10.13 and 2021.10.22.

Debug Messages Revealing Unnecessary Information vulnerability in TLA Media GTM Kit gtm-kit allows Retrieve Embedded Sensitive Data.This issue affects GTM Kit: from n/a through <= 2.4.0.

Apache Airflow versions before 2.10.3 contain a vulnerability that could expose sensitive configuration variables in task logs. This vulnerability allows DAG authors to unintentionally or intentionally log sensitive configuration variables. Unauthorized users could access these logs, potentially exposing critical data that could be exploited to compromise the security of the Airflow deployment. In version 2.10.3, secrets are now masked in task logs to prevent sensitive configuration variables from being exposed in the logging output. Users should upgrade to Airflow 2.10.3 or the latest version to eliminate this vulnerability. If you suspect that DAG authors could have logged the secret values to the logs and that your logs are not additionally protected, it is also recommended that you update those secrets.

C300 information leak due to an analysis feature which allows extracting more memory over the network than required by the function. Honeywell recommends updating to the most recent version of the product. See Honeywell Security Notification for recommendations on upgrading and versioning.

This vulnerability exists in Meon KYC solutions due to debug mode is enabled in certain API endpoints. A remote attacker could exploit this vulnerability by accessing certain unauthorized API endpoints leading to detailed error messages as response leading to disclosure of system related information.

A flaw was found in the Ansible Automation Platform's Event-Driven Ansible. In configurations where verbosity is set to "debug", inventory passwords are exposed in plain text when starting a rulebook activation. This issue exists for any "debug" action in a rulebook and also affects Event Streams.

Advantech WebAccess version 9.1.3 contains an exposure of sensitive information to an unauthorized actor vulnerability that could leak user credentials.

Inappropriate implementation in Passkeys in Google Chrome prior to 140.0.7339.80 allowed a local attacker to obtain potentially sensitive information via debug logs. (Chromium security severity: Low)

A debug messages revealing unnecessary information vulnerability in Fortinet FortiExtender 7.6.0 through 7.6.1, FortiExtender 7.4.0 through 7.4.6, FortiExtender 7.2 all versions, FortiExtender 7.0 all versions may allow an authenticated user to obtain administrator credentials via debug log commands.

Improper sanitization of path in default RouteNotFoundError view in com.vaadin:flow-server versions 1.0.0 through 1.0.14 (Vaadin 10.0.0 through 10.0.18), 1.1.0 prior to 2.0.0 (Vaadin 11 prior to 14), 2.0.0 through 2.6.1 (Vaadin 14.0.0 through 14.6.1), and 3.0.0 through 6.0.9 (Vaadin 15.0.0 through 19.0.8) allows network attacker to enumerate all available routes via crafted HTTP request when application is running in production mode and no custom handler for NotFoundException is provided.

The dormakaba registration units 9002 (PIN Pad Units) have an exposed UART header on the backside. The PIN pad is sending every button press to the UART interface. An attacker can use the interface to exfiltrate PINs. As the devices are explicitly built as Plug-and-Play to be easily replaced, an attacker is easily able to remove the device, install a hardware implant which connects to the UART and exfiltrates the data exposed via UART to another system (e.g. via WiFi).

A vulnerability was found in the OAuth-server. OAuth-server logs the OAuth2 client secret when the logLevel is Debug higher for OIDC/GitHub/GitLab/Google IDPs login options.

Admin cookies are written in clear-text in logs. An attacker can retrieve them and bypass the authentication mechanism. As for the affected products/models/versions, see the reference URL.

Medical Informatics Engineering Enterprise Health includes the user's current session token in debug output. An attacker could convince a user to send this output to the attacker, thus allowing the attacker to impersonate that user. This issue is fixed as of 2025-04-08.

Dell BSAFE SSL-J, versions prior to 6.5, and versions 7.0 and 7.1 contain a debug message revealing unnecessary information vulnerability. This may lead to disclosing sensitive information to a locally privileged user.

Dell BSAFE SSL-J, versions before 6.5 and version 7.0 contain a debug message revealing unnecessary information vulnerability. This may lead to disclosing sensitive information to a locally privileged user. .

An information disclosure vulnerability in Widevine TA log prior to SMR Oct-2021 Release 1 allows attackers to bypass the ASLR protection mechanism in TEE.

In DA, there is a possible out of bounds read due to a missing bounds check. This could lead to local information disclosure, if an attacker has physical access to the device, if a malicious actor has already obtained the System privilege. User interaction is needed for exploitation. Patch ID: ALPS09291146; Issue ID: MSV-2056.

An issue has been discovered in GitLab CE/EE affecting all versions from 17.9 before 17.9.6, and 17.10 before 17.10.4. The runtime profiling data of a specific service was accessible to unauthenticated users.