VAITP

Mistune is a Python Markdown parser with renderers and plugins. In 3.2.0 and earlier, in src/mistune/directives/image.py, the render_figure() function concatenates figclass and figwidth options directly into HTML attributes without escaping. This allows attribute injection and XSS even when HTMLRenderer(escape=True) is used, because these values bypass the inline renderer. Version 3.2.1 contains a patch.

eml_parser serves as a python module for parsing eml files and returning various information found in the e-mail as well as computed information. Prior to 3.0.1, EmlParser.get_raw_body_text() recurses unconditionally for every nested message/rfc822 attachment without any depth limit. An attacker who can supply a badly crafted EML file with approximately 120 nested message/rfc822 parts triggers an unhandled RecursionError and aborts parsing of the message. A 12 KB EML file is enough to crash a worker. Though this causes the parser to crash, it is an unlikely scenario as the suggested EML that crashes the parser would not pass basic RFC compliance tests. This vulnerability is fixed in 3.0.1.

LangChain is a framework for building agents and LLM-powered applications. Prior to 0.3.85 and 1.3.3, LangChain contains older runtime code paths that deserialize run inputs, run outputs, or other application-controlled payloads using overly broad object allowlists. These paths may call load() with allowed_objects="all". This does not enable arbitrary Python object deserialization, but it does allow any trusted LangChain-serializable object to be revived, which is broader than these runtime paths require. As a result, attacker-supplied LangChain serialized constructor dictionaries may cause trusted runtime paths to instantiate classes with untrusted constructor arguments. This vulnerability is fixed in 0.3.85 and 1.3.3.

Mistune is a Python Markdown parser with renderers and plugins. Prior to 3.2.1, the mistune math plugin renders inline math ($...$) and block math ($$...$$) by concatenating the raw user-supplied content directly into the HTML output without any HTML escaping. This occurs even when the parser is explicitly created with escape=True, which is supposed to guarantee that all user-controlled text is sanitised before reaching the DOM. This vulnerability is fixed in 3.2.1.

Lumiverse is a full-featured AI chat application. Prior to 0.9.7, the MCP server creation endpoint validates the command field against an allowlist of binary names but forwards the args array to the child process without any validation. Every binary on the allowlist accepts an inline-code execution flag (-e for node/bun, -c for python3/deno), giving any logged-in user arbitrary OS-level code execution on the Lumiverse server. The route requires only requireAuth (not requireOwner). The server binds on all interfaces (::) and the host-header rebinding check is bypassed trivially by any HTTP client that sends Host: localhost:<port> directly, making this exploitable from any machine with network access to the server port. This vulnerability is fixed in 0.9.7.

Bugsink is a self-hosted error tracking tool. Prior to 2.1.3, Bugsink’s webhook URL validation could be (partially) bypassed because of a mismatch in URL parsing. The original validation logic parsed webhook URLs with Python’s urllib.parse.urlparse, then sent the request with requests.post. For malformed inputs involving backslashes and @, those components can disagree about where the authority ends and which hostname is the real target. A URL may therefore appear to target an allowlisted public hostname during validation, while the HTTP client actually connects to a different host. This vulnerability is fixed in 2.1.3.

A critical remote code execution vulnerability exists in all versions of the HuggingFace transformers library prior to version 5.3.0. The vulnerability allows an attacker to craft a malicious `config.json` file containing the `_attn_implementation_internal` field set to an attacker-controlled HuggingFace Hub repository ID. When a victim loads this model using the standard `AutoModelForCausalLM.from_pretrained()` API, the library downloads and executes arbitrary Python code from the attacker's repository with the victim's full OS privileges. This issue arises due to unfiltered deserialization of configuration attributes, insufficient sanitization of internal fields, and unsandboxed execution of downloaded kernels. The vulnerability bypasses the `trust_remote_code` security mechanism, is invisible to the victim, and exploits the standard documented usage pattern, making it particularly severe. Users are advised to upgrade to version 5.3.0 or later to mitigate this issue.

The MLX inference backend in Docker Model Runner on macOS uses the MLX-LM library, which unconditionally imports and executes arbitrary Python files from model directories via the model_file configuration field in config.json. When a model's config.json specifies a model_file pointing to a Python file, MLX-LM uses importlib to load and execute it with no trust_remote_code gate or equivalent safety check. The MLX backend runs without sandboxing, resulting in arbitrary code execution on the Docker host as the Docker Desktop user. Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model from an attacker-controlled OCI registry and request inference.

The vllm-metal inference backend in Docker Model Runner on macOS unconditionally sets trust_remote_code=True when loading model tokenizers, and runs without sandboxing. This causes transformers.AutoTokenizer.from_pretrained() to import and execute arbitrary Python files included in any model pulled from an OCI registry, resulting in arbitrary code execution on the Docker host as the Docker Desktop user when inference is triggered. Any container on the Docker network can trigger this by calling the model-runner.docker.internal API to pull a malicious model and request inference.

BentoML is a Python library for building online serving systems optimized for AI apps and model inference. In versions 1.4.38 and prior, the build packaging workflow follows attacker-controlled symlinks inside the build context and copies the referenced file contents into the generated Bento artifact. If a victim builds an untrusted repository or other attacker-supplied build context, the attacker can place a symlink such as loot.txt -> /tmp/outside-marker.txt or a link to a more sensitive local file. When bentoml build runs, BentoML dereferences the symlink and packages the target file contents into the Bento. The leaked file can then propagate further through export, push, or containerization workflows. An attacker can exfiltrate local files from the build host into the Bento artifact, exposing secrets such as cloud credentials, SSH keys, API tokens, environment files, or other sensitive local configurations. Because Bento artifacts are commonly exported, uploaded, stored, or containerized after build, the leaked file contents can spread beyond the original build machine. This issue has been fixed in version 1.4.39.