Lundi 12 octobre 2026

picklescan

picklescan before 0.0.29 fails to detect the built-in python profile.Profile.run function when used in pickle reduce methods, allowing attackers to execute arbitrary code. Remote attackers can craft malicious pickle files that bypass picklescan detection and achieve code execution upon deserialization.

picklescan

picklescan before 0.0.29 fails to detect the built-in Python trace.Trace.runctx function when used in pickle file reduce methods, allowing attackers to execute arbitrary code. Remote attackers can craft malicious pickle files with trace.Trace.runctx payloads that bypass picklescan detection and execute code upon pickle.load() invocation.

picklescan

picklescan through 0.0.26 fails to detect malicious pickle files that invoke idlelib.pyshell.ModifiedInterpreter.runcode in __reduce__ methods. Attackers can embed undetected code in pickle files that executes arbitrary commands when the file is loaded via pickle.load(), enabling supply chain attacks on PyTorch models and saved Python objects. This is fixed in version 0.0.30.

Picklescan

Picklescan before 0.0.33 fails to detect the numpy.f2py.crackfortran.getlincoef gadget in pickle __reduce__ methods, allowing arbitrary code execution. Attackers can craft malicious pickle files that execute arbitrary Python code when loaded, bypassing Picklescan's safety checks and enabling supply-chain poisoning of shared model files.

EmilStenstrom justhtml

justhtml through 1.9.1 (fixed in 1.10.0) is vulnerable to uncontrolled recursion leading to denial of service. During JustHTML() construction, TreeBuilder.finish() unconditionally calls _populate_selectedcontent(), which recursively traverses the DOM tree via _find_elements()/_find_element() without a depth bound. An attacker who can supply HTML for parsing can provide deeply nested elements (e.g., ~1000 nested <div> tags, roughly 11 KB) to exceed CPython's default recursion limit and trigger an unhandled RecursionError, which may abort parsing, fail requests, or terminate a worker/process depending on the host application's exception handling.

facebookresearch hydra

Hydra is a framework for elegantly configuring complex applications. Prior to 1.3.4, hydra.utils.instantiate() resolves and calls Python objects selected by configuration through _resolve_target() in hydra/_internal/instantiate/_instantiate2.py, allowing attacker-controlled target values and arguments to choose dangerous callables. A consuming application, library, CLI workflow, or model loader that passes untrusted configuration, CLI overrides, or model metadata into hydra.utils.instantiate() can therefore execute arbitrary code in its own process, including reading or modifying files and credentials or terminating the process. Version 1.3.4 adds target blocking with an explicit HYDRA_INSTANTIATE_ALLOWLIST_OVERRIDE escape hatch. This issue is fixed in version 1.3.4.

gitpython-developers GitPython · Gitpython project gitpython

GitPython before 3.1.59 omits --separate-git-dir from unsafe_git_clone_options, allowing attackers to create arbitrary git directories outside the intended clone destination. Attackers can pass a separate_git_dir parameter to Repo.clone_from() or Repo.clone() to redirect repository metadata to an attacker-controlled filesystem path, enabling arbitrary directory creation and potential hook execution.

gitpython-developers GitPython · Gitpython project gitpython

GitPython before 3.1.59 fails to disable merge_includes when parsing .gitmodules, allowing attackers to disclose local file content by including arbitrary file paths via [include] directives. Attackers can craft a malicious .gitmodules file with include directives pointing to sensitive files; when repo.submodules is accessed, GitConfigParser raises MissingSectionHeaderError embedding the target file's first line verbatim in the exception message.

lepture mistune

Mistune is a Python Markdown parser with renderers and plugins. Versions 3.3.0 through 3.3.2 are vulnerable to DoS through deeply nested tokens. HTML rendering creates deeply nested emphasis tokens from consecutive asterisk characters, and recursive rendering in HTMLRenderer.render_token() can exceed Python's recursion limit and raise RecursionError, allowing crafted Markdown to crash a parsing process. This issue is fixed in version 3.3.3

flairNLP flair

The official Flair wheels for 0.15.0 and 0.15.1 still contain flair/models/clustering.py, whose ClusteringModel.load static method returns pickle.loads(joblib.load(str(model_file))) and so executes arbitrary Python while loading a model file. Loading a model supplied by an attacker therefore runs that attacker's code with the privileges of the loading process. This is the same sink and the same file as CVE-2024-10073, which records 0.15.0 as the fixed version on the basis that clustering support was dropped in that release; the module was removed from the documented API but remains present in the distributed artifact and reachable by importing flair.models.clustering directly, so the earlier record's fixed version does not hold for the shipped package.

xorbitsai inference

Xinference loads models with Hugging Face remote code execution unconditionally enabled, and before version 2.12.0 exposes no setting to disable it. Six loader call sites pass trust_remote_code=True as a literal or as an unconditional default: RerankModel._get_tokenizer in xinference/model/rerank/core.py, SentenceTransformerRerankModel.load in xinference/model/rerank/sentence_transformers/core.py, SentenceTransformerEmbeddingModel.load in xinference/model/embedding/sentence_transformers/core.py, FlagEmbeddingModel.load in xinference/model/embedding/flag/core.py, and two sites in xinference/model/llm/transformers/core.py where PytorchModel._sanitize_model_config and PytorchModel._get_components default the value to True. Because a caller with model launch access can register a model whose type is unknown and supply an arbitrary model path, the server reaches _auto_detect_type and then AutoTokenizer.from_pretrained, which imports and executes Python declared by the model directory's own tokenizer_config.json auto_map, running attacker-supplied code with the privileges of the worker process. Version 2.12.0 gates every site behind allow_trust_remote_code and the XINFERENCE_TRUST_REMOTE_CODE setting, permitting remote code only for bundled built-in models.