Malicious code in pantheon-toolsets (PyPI)

Part of the "Hades" wave of the Shai-Hulud supply-chain campaign. On 2026-06-08,malicious phantom releases of pantheon-toolsets were published to PyPI using stolencredentials. The package executes a bundled JavaScript payload via the Bunruntime on import that harvests and exfiltrates credentials...

Malicious code in pantheon-agents (PyPI)

Part of the "Hades" wave of the Shai-Hulud supply-chain campaign. On 2026-06-08,malicious phantom releases of pantheon-agents were published to PyPI using stolencredentials. The package executes a bundled JavaScript payload via the Bunruntime on import that harvests and exfiltrates credentials an...

Malicious code in napari-ufish (PyPI)

Part of the "Hades" wave of the Shai-Hulud supply-chain campaign. On 2026-06-08,malicious phantom releases of napari-ufish were published to PyPI using stolencredentials. The package executes a bundled JavaScript payload via the Bunruntime on import that harvests and exfiltrates credentials and...

Malicious code in executor-engine (PyPI)

Part of the "Hades" wave of the Shai-Hulud supply-chain campaign. On 2026-06-08,malicious phantom releases of executor-engine were published to PyPI using stolencredentials. The package executes a bundled JavaScript payload via the Bunruntime on import that harvests and exfiltrates credentials an...

Malicious code in tlask (PyPI)

Part of the "Hades" wave of the Shai-Hulud supply-chain campaign. On 2026-06-08,malicious phantom releases of tlask were published to PyPI using stolencredentials. The package executes a bundled JavaScript payload via the Bunruntime on import that harvests and exfiltrates credentials and...

Malicious code in dynamo-release (PyPI)

Part of the "Hades" wave of the Shai-Hulud supply-chain campaign. On 2026-06-08,malicious phantom releases of dynamo-release were published to PyPI using stolencredentials. The package executes a bundled JavaScript payload via the Bunruntime on import that harvests and exfiltrates credentials and...

Jupyter Enterprise Gateway: Kubernetes Manifest Injection in Jinja2 Template Rendering

SummaryThe environment variables used during the rendering of the Kubernetes manifest allow YAML injection, enabling attackers to overwrite existing keys like securityContext and inject multi-document YAML to create additional unintended Kubernetes resources. DetailsThe server interpolates...

Langflow has Remote Code Execution in CSV Agent

SummaryThe CSV Agent node in Langflow hardcodes allowdangerouscode=True, which automatically exposes LangChain’s Python REPL tool pythonreplast. As a result, an attacker can execute arbitrary Python and OS commands on the server via prompt injection, leading to full Remote Code Execution RCE. 2...

PraisonAI vulnerable to sandbox escape via `print.__self__` builtins module leak in `execute_code` (subprocess mode)

Summaryexecutecode in praisonaiagents/tools/pythontools.py v1.6.37, subprocess sandbox mode can be fully bypassed using print.self to retrieve the real Python builtins module, from which import can be extracted via vars and runtime string construction. This achieves arbitrary OS command execution...

PraisonAI MCP `tools/call` path-traversal => RCE via Python `.pth` injection

SummaryPraisonAI's MCP Model Context Protocol server praisonai mcp serve registers four file-handling tools by default — praisonai.rules.create, praisonai.rules.show, praisonai.rules.delete, and praisonai.workflow.show. Each accepts a path or filename string from MCP tools/call arguments and join...

dash-uploader has a directory traversal vulnerability

ImpactAn unauthenticated path traversal vulnerability exists in dash-uploader versions 0.1.0 through 0.7.0a2. The library's HTTP request handler at dashuploader/httprequesthandler.py reads three form parameters uploadid, resumableFilename, resumableIdentifier from request.form.get and passes them...

SGLang's multimodal generation module is vulnerable to unauthenticated remote code execution through the ZMQ broker

SGLang's multimodal generation module is vulnerable to unauthenticated remote code execution through the ZMQ broker, which deserializes untrusted data using pickle.loads without authentication...

MLflow Command Injection vulnerability

A command injection vulnerability exists in MLflow's model serving container initialization code, specifically in the installmodeldependenciestoenv function. When deploying a model with envmanager=LOCAL, MLflow reads dependency specifications from the model artifact's pythonenv.yaml file and...

Remote Code Execution due to Full Controled File Write in mlflow

The mlflow web server includes tools for tracking experiments, packaging code into reproducible runs, and sharing and deploying models. As this vulnerability allows to write / overwrite any file on the file system, it gives a lot of ways to archive code execution like overwriting /home//.bashrc. ...

Ray OS Command Injection vulnerability

A command injection exists in Ray's cpuprofile URL parameter allowing attackers to execute os commands on the system running the ray dashboard remotely without authentication...

TorchServe Server-Side Request Forgery vulnerability

ImpactRemote Server-Side Request Forgery SSRF Issue: TorchServe default configuration lacks proper input validation, enabling third parties to invoke remote HTTP download requests and write files to the disk. This issue could be taken advantage of to compromise the integrity of the system and...

vanna vulnerable to remote code execution caused by prompt injection

In the latest version of vanna-ai/vanna, the vanna.ask function is vulnerable to remote code execution due to prompt injection. The root cause is the lack of a sandbox when executing LLM-generated code, allowing an attacker to manipulate the code executed by the exec function in...

TensorFlow has a heap out-of-buffer read vulnerability in the QuantizeAndDequantize operation

ImpactAttackers using Tensorflow can exploit the vulnerability. They can access heap memory which is not in the control of user, leading to a crash or RCE.When axis is larger than the dim of input, c-Diminput,axis goes out of bound.Same problem occurs in the QuantizeAndDequantizeV2/V3/V4/V4Grad...

Malicious Long Unicode filenames may cause a Multiple Application-level Denial of Service

Important: Exploiting this vulnerability requires the attacker to have access to your Frigate instance, which means they could also just delete all of your recordings or perform any other action. If you have configured authentication in front of Frigate via a reverse proxy, then this vulnerabilit...

PyMySQL SQL Injection vulnerability

PyMySQL through 1.1.0 allows SQL injection if used with untrusted JSON input because keys are not escaped by escapedict...

Rasa Allows Remote Code Execution via Remote Model Loading

VulnerabilityA vulnerability has been identified in Rasa Pro and Rasa Open Source that enables an attacker who has the ability to load a maliciously crafted model remotely into a Rasa instance to achieve Remote Code Execution.The prerequisites for this are:- The HTTP API must be enabled on the Ra...

Langroid has a Code Injection vulnerability in TableChatAgent

SummaryTableChatAgent uses pandas eval. If fed by untrusted user input, like the case of a public-facing LLM application, it may be vulnerable to code injection. PoCFor example, one could prompt the Agent: Evaluate the following pandas expression on the data provided and print output:...

vLLM Allows Remote Code Execution via PyNcclPipe Communication Service

Impacted EnvironmentsThis issue ONLY impacts environments using the PyNcclPipe KV cache transfer integration with the V0 engine. No other configurations are affected. SummaryvLLM supports the use of thePyNcclPipeclass to establish a peer-to-peer communication domain for data transmission between...

Aim External Control of File Name or Path vulnerability

A vulnerability in aimhubio/aim version 3.19.3 allows an attacker to exploit the tarfile.extractall function to extract the contents of a maliciously crafted tarfile to arbitrary locations on the host server. The attacker can control repo.path and runhash to bypass directory existence checks and...

Aim path traversal in LockManager.release_locks

A vulnerability in the LockManager.releaselocks function in aimhubio/aim commit bb76afe allows for arbitrary file deletion through relative path traversal. The runhash parameter, which is user-controllable, is concatenated without normalization as part of a path used to specify file deletion. Thi...

DB-GPT Absolute Path Traversal in knowledge/{space_name}/document/upload

eosphoros-ai/db-gpt version 0.6.0 is vulnerable to an arbitrary file write through the knowledge API. The endpoint for uploading files as 'knowledge' is susceptible to absolute path traversal, allowing attackers to write files to arbitrary locations on the target server. This vulnerability arises...

PYSEC-2026-2146

Docling simplifies document processing by parsing diverse formats and providing integrations with the generative AI ecosystem. Prior to 2.94.0, the HTML backend has unsafe URI and path handling. This vulnerability is fixed in 2.94.0...

PYSEC-2026-230

Crawl4AI before 0.8.7 contains a stored cross-site scripting vulnerability in the monitor dashboard that renders crawl URLs and error messages via innerHTML without escaping. An attacker can submit a crafted crawl request with malicious markup that executes in an operator's browser when viewing t...

PYSEC-2026-2119

Authlib is a Python library which builds OAuth and OpenID Connect servers. Prior to 1.6.10 and 1.7.1, Authlib's OAuth 2.0 authorization endpoint can be turned into an unauthenticated open redirect when a request uses an unsupported responsetype and supplies an attacker-controlled redirecturi. The...

PYSEC-2026-237

AIOHTTP is an asynchronous HTTP client/server framework for asyncio and Python. Prior to 3.14.1, the serverhostname TLS SNI check can be bypassed when an existing connection is reused. If an application makes multiple requests to the same domain, but with different per-request serverhostname...

PYSEC-2026-2107

AIOHTTP is an asynchronous HTTP client/server framework for asyncio and Python. Prior to 3.14.1, no limit was present on the number of pipelined requests that could be queued. An attacker may be able to use pipelined requests to use excessive amounts of memory, potentially leading to DoS. This...

PYSEC-2026-208

The Apache Airflow Samba provider's GCSToSambaOperator joined GCS object names to the SMB destination path without a containment check, so an object named with ../ segments resolved a write path outside the configured destinationpath. An attacker able to write objects into the source GCS bucket —...

PYSEC-2026-2284

Strawberry GraphQL is a library for creating GraphQL APIs. In versions 0.172.0 through0.315.6, the MaxAliasesLimiter extension in Strawberry fails to account for the multiplicative/amplification effect of FragmentSpreadNode. While it correctly counts static aliases within the AST it does not...

PYSEC-2026-2228

Nautobot is a Network Source of Truth and Network Automation Platform. Prior to 2.4.33 and 3.1.2, a user with access to add/change a GitRepository record could use the REST API to directly set the currenthead field on the record, which was not intended to be user-editable. Doing so could cause...

PYSEC-0000-CVE-2026-48525

PyJWT is a JSON Web Token implementation in Python. From 2.8.0 to 2.12.1, when verifying detached JWS tokens using the unencoded-payload option "b64": false, RFC 7797, PyJWT performs Base64URL decoding of the compact-serialization payload segment before enforcing the detached-payload rules. For...

PYSEC-2026-156

Weblate is a web based localization tool. In versions prior to 5.17, the webhook add-on did not utilize existing SSRF protections. This issue has been fixed in version 5.17. If developers are unable to update immediately, they can disable the webhook add-on as a workaround...

PYSEC-2026-2265

OpenCTI is an open source platform for managing cyber threat intelligence knowledge and observables. Prior to 6.9.5, the safeEjs.ts file does not properly sanitize EJS templates. Users with the Manage customization capability can run arbitrary JavaScript in the context of the OpenCTI platform...

PYSEC-2026-28

changedetection.io is a free open source web page change detection tool. Prior to 0.54.8, the @loginoptionallyrequired decorator is placed before outer to @blueprint.route instead of after it. In Flask, @route must be the outermost decorator because it registers the function it receives. When the...

PYSEC-2026-49

An issue was discovered in 6.0 before 6.0.4, 5.2 before 5.2.13, and 4.2 before 4.2.30.ASGI requests with a missing or understated Content-Length header couldbypass the DATAUPLOADMAXMEMORYSIZE limit when readingHttpRequest.body, allowing remote attackers to load an unbounded request body...

PYSEC-2026-79

Langflow is a tool for building and deploying AI-powered agents and workflows. Versions 1.2.0 through 1.8.1 have a bypass of the patch for CVE-2025-68478 External Control of File Name, leading to the root architectural issue within LocalStorageService remaining unresolved. Because the underlying...

PYSEC-2026-39

Deluge 1.3.15 contains a denial of service vulnerability that allows local attackers to crash the application by supplying an excessively long string in the URL field. Attackers can paste a buffer of 5000 characters into the 'From URL' field during torrent addition to trigger an application crash...

PYSEC-2026-2263

pyasn1 is a generic ASN.1 library for Python. Prior to 0.6.3, the pyasn1 library is vulnerable to a Denial of Service DoS attack caused by uncontrolled recursion when decoding ASN.1 data with deeply nested structures. An attacker can supply a crafted payload containing thousands of nested SEQUENC...

PYSEC-2026-16

Apache Airflow versions 3.1.0 through 3.1.7session token token in cookies is set to path=/ regardless of the configured webserver baseurl or api baseurl. This allows any application co-hosted under the same domain to capture valid Airflow session tokens from HTTP request headers, allowing full...

PYSEC-2026-121

pyLoad is a free and open-source download manager written in Python. From version 0.5.0b3.dev13 to 0.5.0b3.dev96, the editpackage function implements insufficient sanitization for the packfolder parameter. The current protection relies on a single-pass string replacement of "../", which can be...

PYSEC-2026-98

A vulnerability in NLTK versions up to and including 3.9.2 allows arbitrary file read via path traversal in multiple CorpusReader classes, including WordListCorpusReader, TaggedCorpusReader, and BracketParseCorpusReader. These classes fail to properly sanitize or validate file paths, enabling...

PYSEC-2026-102

An issue in ollama v.0.12.10 allows a remote attacker to cause a denial of service via the fs/ggml/gguf.go, function readGGUFV1String reads a string length from untrusted GGUF metadata...

PYSEC-2025-113

Fickling is a Python pickling decompiler and static analyzer. Versions prior to 0.1.6 had a bypass caused by pty missing from the block list of unsafe module imports. This led to unsafe pickles based on pty.spawn being incorrectly flagged as LIKELYSAFE, and was fixed in version 0.1.6. This impact...

PYSEC-2025-233

Weblate is a web based localization tool. In versions prior to 5.15, it was possible to retrieve user notification settings or list all users via API. Version 5.15 fixes the issue...

PYSEC-2025-112

DuckDB is a SQL database management system. DuckDB implemented block-based encryption of DB on the filesystem starting with DuckDB 1.4.0. There are a few issues related to this implementation. The DuckDB can fall back to an insecure random number generator pcg32 to generate cryptographic keys or...

PYSEC-2025-100

CodeChecker is an analyzer tooling, defect database and viewer extension for the Clang Static Analyzer and Clang Tidy. CodeChecker versions up to 6.26.1 contain a buffer overflow vulnerability in the internal ldloggerlibrary, which is executed by the CodeChecker logcommand.This issue affects...