CVE-2026-95280

A flaw was found in chromium-browser. Arbitrary code execution within the browser's restricted security environment sandbox is possible due to a race condition in the V8 JavaScript engine. A remote attacker can exploit this vulnerability by enticing a user to visit a specially crafted web page...

HARDE

HARDE: ランタイムリスク検出と実行制御のためのエージェントハーネス最適化 HARDE: Optimizing Agent Harnesses for Runtime Risk Detection and Execution Control のコードリリース。 このリポジトリには、2つの相補的なエントリポイントファミリーが含まれています： domains/: Meta-Harness ベースラインと、各ベンチマーク向け HARDE の2つのステージ。 personalizedharness/:...

CVE-2026-47187

A flaw was found in SSHFS. A malicious SFTP Secure File Transfer Protocol server can supply crafted symbolic links containing absolute paths or parent directory traversal sequences that are resolved by the client system. When a local user or automated tool accesses or copies files through the...

Trestle SSTI in Jinja2 include tags allows arbitrary code execution (Incomplete fix of CVE-2026-46439)

Reporter: Cavan Loughran, Celvex Group Inc.Summary-------The fix for CVE-2026-46439 3.12.2 / 4.0.3 removed the recursive re-render loop in trestle/core/commands/author/jinja.py rendertemplate, but the custom include tags in trestle/core/jinja/tags.py MDSectionInclude, MDCleanInclude still re-pars...

Decepticon: Role-boundary forgery via ChatML special-token literals in web crawl output composed into LLM context

SummaryDecepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK Bring Your Own Key deployment model, users configure their own LLM credentials to any OpenAI-compatible...

Decepticon: Role-boundary forgery via ChatML special-token literals in web crawl output composed into LLM context

SummaryDecepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK Bring Your Own Key deployment model, users configure their own LLM credentials to any OpenAI-compatible...

plone.app.portlets Vulnerable to Remote Code Execution via TALES Injection

ImpactThe Classic portlet plone.app.portlets.portlets.classic used its user-supplied template/macro fields to build a TALES path expression that was then evaluated by the TAL path helper. Because the value was interpreted as a full TALES expression, a user able to add or edit a Classic portlet...

Decepticon: Role-boundary forgery via ChatML special-token literals in web crawl output composed into LLM context

SummaryDecepticon wraps web crawl results — the output of agent reconnaissance against target services — into LLM messages without neutralizing ChatML special-token literals. Under the BYOK Bring Your Own Key deployment model, users configure their own LLM credentials to any OpenAI-compatible...

LMDeploy vulnerable to arbitrary code execution via eval() of untrusted quant_dtype in model config loading

Summarylmdeploy = latest contains a code injection vulnerability in lmdeploy/pytorch/config.py line 620 that allows an attacker to execute arbitrary Python code by publishing a malicious HuggingFace model with a crafted quantizationconfig.quantdtype value. When a user loads the model with lmdeplo...

LMDeploy has Remote Code Execution by Pickle Deserialization via handle_zmq_recv in lmdeploy/lmdeploy/pytorch/disagg/conn/engine_conn.py

SummaryLMDeploy's PyTorch DistServe/PD-disaggregation control plane usedrecvpyobj to deserialize messages received through a ZeroMQ PULLsocket. PyZMQ implements recvpyobj using Python pickledeserialization, which can execute arbitrary code while reconstructingan object.The peer address used by th...

RestrictedPython vulnerable to sandbox escape via string.Formatter field resolution

ImpactRestrictedPython could allow a sandbox escape when a policy exposes the standard library string module, or otherwise exposes string.Formatter, to restricted code.string.Formatter field resolution methods such as getfield can perform attribute and item traversal internally and return live...

LMdeploy has Remote Code Execution by Pickle Deserialization via zmq_rpc.call_and_response() in InterLM/lmdeploy

DescriptionThe LMdeployimplements anrpc server AsyncRPCServer in zmqrpc.pyfor supporting the RPC communications. In its core functionalitycallandresponse, I found it will directly use thepickles.loadsto deserialize the received messages without any sanitization, hence resulting in a remote code...

MySQL MCP Server: Missing Origin/Host Validation in SSE Transport Enables Unauthenticated SQL Execution (DNS Rebinding / Direct Exposure)

SummaryIn SSE/HTTP transport mode, mysqlmcpserver constructs SseServerTransport without passing securitysettings. As a result, the MCP Python SDK's DNS-rebinding protection Origin/Host header validation is disabled; the Starlette application has no CORS or TrustedHost middleware; and the service...

Django GeoDjango spatial lookups allow file writes and outbound requests through GDAL raster parsing

An issue was discovered in Django 5.2 before 5.2.17 and 6.0 before 6.0.8.GeoDjango spatial lookups optimistically parse the right-hand-side value as a raster by passing it to the django.contrib.gis.gdal.GDALRaster constructor. Any value used in a spatial lookup against a GeometryField or...

sentence-transformers local model loading bypasses trust_remote_code and executes custom Python

sentence-transformers contains a security control bypass vulnerability that allows attackers to achieve arbitrary code execution by exploiting a logic flaw in the importmoduleclass helper within sentencetransformers/util/misc.py, where the guard condition includes an 'or...

Picklescan has a missing detection when calling built-in python idlelib.autocomplete.AutoComplete.fetch_completions

SummaryUsing idlelib.autocomplete.AutoComplete.fetchcompletions, which is a built-in python library function to execute remote pickle file. DetailsThe attack payload executes in the following steps:First, the attacker craft the payload by calling to...

Picklescan missing detection when calling pytorch function torch.jit.unsupported_tensor_ops.execWrapper

SummaryUsing torch.jit.unsupportedtensorops.execWrapper function, which is a pytorch library function to execute remote pickle file. DetailsThe attack payload executes in the following steps:First, the attacker craft the payload by calling to torch.jit.unsupportedtensorops.execWrapper function in...

Picklescan missing detection when calling built-in python library function timeit.timeit()

SummaryUsing timeit.timeit function, which is a built-in python library function to execute remote pickle file. DetailsPickle’s deserialization process is known to allow execution of function via reduce method. While Picklescan is meant to detect such exploits, this attack evades detection by...

Picklescan has a missing detection when calling built-in python profile.Profile.runctx

SummaryUsing profile.Profile.runctx, which is a built-in python library function to execute remote pickle file. DetailsThe attack payload executes in the following steps:First, the attacker craft the payload by calling to profile.Profile.runctx function in reduce methodThen when the victim after...

CVE-2026-42533

CVE-2026-42533 — nginx map/regex capture clobbering tracking site Source for the CVE-2026-42533 patch-status tracker: a single-page site recording which distributions have shipped a fix for the nginx map/regex capture-clobbering heap overflow. Where the facts live Everything about the bug —...