Open WebUI has Broken Access Control in Tool Valves

Summary Broken Access Control in Tool ValvesOpen WebUI supports function calling through "Tools". Function calling allows an LLM to reliably connect to external tools and interact with external APIs. Exemplary use-cases include connecting to an internal knowledge base, retrieving emails from an...

PraisonAI Has Sandbox Escape via shell=True and Bypassable Blocklist in SubprocessSandbox

SummarySubprocessSandbox in all modes BASIC, STRICT, NETWORKISOLATED calls subprocess.run with shell=True and relies solely on string-pattern matching to block dangerous commands. The blocklist does not include sh or bash as standalone executables, allowing trivial sandbox escape in STRICT mode v...

PraisonAI: SSRF via Unvalidated api_base in passthrough() Fallback

Summarypassthrough and apassthrough in praisonai accept a caller-controlled apibase parameter that is concatenated with endpoint and passed directly to httpx.Client.request when the litellm primary path raises AttributeError. No URL scheme validation, private IP filtering, or domain allowlist is...

PraisonAI: Shell Injection in run_python() via Unescaped $() Substitution

Summaryrunpython in praisonai constructs a shell command string by interpolating user-controlled code into python3 -c "" and passing it to subprocess.run..., shell=True. The escaping logic only handles \ and ", leaving $ and backtick substitutions unescaped, allowing arbitrary OS command executio...

alerta-server has potential SQL Injection vulnerability in Query String Syntax (q=) API

ImpactThe Query string search API q= was vulnerable to SQL injection via the Postgres query parser, which built WHERE clauses by interpolating user-supplied search terms directly into SQL strings via f-strings. PatchesFixed in v9.1.0. The Postgres query parser now uses parameterized queries with...

OpenStack Glance is affected by Server-Side Request Forgery (SSRF)

OpenStack Glance versions = 30.0.0 30.1.1, == 31.0.0 are affected by Server-Side Request Forgery SSRF. By use of HTTP redirects, an authenticated user can bypass URL validation checks and redirect to internal services. Only the glance image import functionality is affected. In particular, the...

PraisonAI Has ReDoS via Unvalidated User-Controlled Regex in MCPToolIndex.search_tools()

SummaryMCPToolIndex.searchtools compiles a caller-supplied string directly as a Python regular expression with no validation, sanitization, or timeout. A crafted regex causes catastrophic backtracking in the re engine, blocking the Python thread for hundreds of seconds and causing a complete...

FastMCP: Missing Consent Verification in OAuth Proxy Callback Facilitates Confused Deputy Vulnerabilities

SummaryWhile testing the GitHubProvider OAuth integration, which allows authentication to a FastMCP MCP server via a FastMCP OAuthProxy using GitHub OAuth, it was discovered that the FastMCP OAuthProxy does not properly validate the user's consent upon receiving the authorization code from GitHub...

Open WebUI's Insecure Direct Object Reference (IDOR) allows access to other users' memories

SummaryAny authenticated user can read other users' private memories via /api/v1/retrieval/query/collection DetailsVulnerability 1: Missing authorization in collection queryingIn backend/openwebui/routers/retrieval.py, the querycollectionhandler function accepts a list of collectionnames but...

Azure Data Explorer MCP Server: KQL Injection in multiple tools allows MCP client to execute arbitrary Kusto queries

Summaryadx-mcp-server ListDictstr, Any: client = getkustoclient query = f"tablename | getschema" ListDictstr, Any: client = getkustoclient query = f"tablename | sample samplesize" ListDictstr, Any: client = getkustoclient query = f".show table tablename details" -- KQL injection resultset =...

Apache Airflow Provider for Databricks: TLS Certificate Verification is Disabled in Databricks Provider K8s Token Exchange

Improper Certificate Validation vulnerability in Apache Airflow Provider for Databricks. Provider code did not validate certificates for connections to Databricks back-end which could result in a man-of-a-middle attack that traffic is intercepted and manipulated or credentials exfiltrated w/o...

python-ecdsa: Denial of Service via improper DER length validation in crafted private keys

SummaryAn issue in the low-level DER parsing functions can cause unexpected exceptions to be raised from the public API functions.1. ecdsa.der.removeoctetstring accepts truncated DER where the encoded length exceeds the available buffer. For example, an OCTET STRING that declares a length of 4096...

Giskard Agents have Server-side template injection via ChatWorkflow.chat() using non-sandboxed Jinja2 Environment

SummaryChatWorkflow.chatmessage passes its string argument directly as a Jinja2 template source to a non-sandboxed Environment. A developer who passes user input to this method enables full remote code execution via Jinja2 class traversal.The method name chat and parameter name message naturally...

Home Assistant has stored XSS in history-graphs

SummaryThe "remaining charge time"-sensor for mobile phones imported/included from Android Auto it appears is vulnerable to the same issue as CVE-2025-62172.This also indicates that any sensor showing their name in the history-graph, is likely to be vulnerable to this issue. DetailsAnother entity...

Home Assistant has stored XSS in Map-card through malicious device name

SummaryAn authenticated party can add a malicious name to their device entity, allowing for Cross-Site Scripting attacks against anyone who can see a dashboard with a Map-card which includes that entity. It requires that the victim hovers over an information point The lines or the dots representi...

Hugging Face Smolagents has an Injection issue

A weakness has been identified in huggingface smolagents 1.25.0.dev0. This affects the function evaluateaugassign/evaluatecall/evaluatewith of the file src/smolagents/localpythonexecutor.py of the component Incomplete Fix CVE-2025-9959. This manipulation causes code injection. It is possible to...

Langflow: Authenticated Users Can Read, Modify, and Delete Any Flow via Missing Ownership Check

Vulnerability IDOR in GET/PATCH/DELETE /api/v1/flow/flowidThe readflow helper in src/backend/base/langflow/api/v1/flows.py branched on the AUTOLOGIN setting to decide whether to filter by userid. When AUTOLOGIN was False i.e., authentication was enabled, neither branch enforced an ownership check...

MLFlow allows Tracing + Assessments Access

In the latest version of mlflow/mlflow, when the basic-auth app is enabled, tracing and assessment endpoints are not protected by permission validators. This allows any authenticated user, including those with NOPERMISSIONS on the experiment, to read trace information and create assessments for...

Langflow: Authenticated Users Can Read, Modify, and Delete Any Flow via Missing Ownership Check

Vulnerability IDOR in GET/PATCH/DELETE /api/v1/flow/flowidThe readflow helper in src/backend/base/langflow/api/v1/flows.py branched on the AUTOLOGIN setting to decide whether to filter by userid. When AUTOLOGIN was False i.e., authentication was enabled, neither branch enforced an ownership check...

Open WebUI's process_files_batch() endpoint missing ownership check, allows unauthorized file overwrite

SummaryAny authenticated user can overwrite any file's content by ID through the POST /api/v1/retrieval/process/files/batch endpoint. The endpoint performs no ownership check, so a regular user with read access to a shared knowledge base can obtain file UUIDs via GET /api/v1/knowledge/id/files an...

Open WebUI vulnerable to Path Traversal in `POST /api/v1/audio/transcriptions`

SummaryAn unsanitised filename field in the speech-to-text transcription endpoint allows any authenticated non-admin user to trigger a FileNotFoundError whose message — including the server's absolute DATADIR path — is returned verbatim in the HTTP 400 response body, confirming information...

pypdf: Possible infinite loop during recovery attempts in DictionaryObject.read_from_stream

ImpactAn attacker who uses this vulnerability can craft a PDF which leads to an infinite loop. This requires reading a file in non-strict mode. PatchesThis has been fixed in pypdf==6.9.2. WorkaroundsIf users cannot upgrade yet, consider applying the changes from PR 3693...

Pygments has Regular Expression Denial of Service (ReDoS) due to Inefficient Regex for GUID Matching

A security flaw has been discovered in pygments before 2.20.0. The impacted element is the function AdlLexer of the file pygments/lexers/archetype.py. The manipulation results in inefficient regular expression complexity. The attack is only possible with local access. The exploit has been release...

NVIDIA NeMo Framework contains a vulnerability leading to Remote Code Execution

NVIDIA NeMo Framework contains a vulnerability where an attacker may cause remote code execution. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, information disclosure and data tampering...

NVIDIA NeMo Framework contains an RCE vulnerability in checkpoint loading

NVIDIA NeMo Framework contains a vulnerability in checkpoint loading where an attacker could cause remote code execution. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, information disclosure and data tampering...

MindSQL is vulnerable to Code Injection through its ask_db function

A vulnerability was found in Mindinventory MindSQL up to 0.2.1. Impacted is the function askdb of the file mindsql/core/mindsqlcore.py. Performing a manipulation results in code injection. The attack can be initiated remotely. The exploit has been made public and could be used. The vendor was...

MobSF has SQL Injection in its SQLite Database Viewer Utils

DescriptionMobSF's readsqlite function in mobsf/MobSF/utils.py lines 542-566 uses Python string formatting % to construct SQL queries with table names read from a SQLite database's sqlitemaster table. When a security analyst uses MobSF to analyze a malicious mobile application containing a crafte...

Open WebUI has unauthorized deletion of knowledge files

SummaryAn access control check is missing when deleting a file from a knowledge base. The only check being done is that the user has write access to the knowledge base or is admin, but NOT that the file actually belongs to this knowledge base. It is thus possible to delete arbitrary files from...

pyLoad SETTINGS Permission Users Can Achieve Remote Code Execution via Unrestricted Reconnect Script Configuration

SummaryThe setconfigvalue API endpoint allows users with the non-admin SETTINGS permission to modify any configuration option without restriction. The reconnect.script config option controls a file path that is passed directly to subprocess.run in the thread manager's reconnect logic. A SETTINGS...

Vanna has a SQL injection in the remove_training_data function

A flaw has been found in vanna-ai vanna up to 2.0.2. This impacts the function removetrainingdata of the file src/vanna/legacy/google/bigqueryvector.py. This manipulation of the argument ID causes sql injection. The attack can be initiated remotely. The exploit has been published and may be used...

DeepDiff has Memory Exhaustion DoS through SAFE_TO_IMPORT

SummaryThe pickle unpickler RestrictedUnpickler validates which classes can be loaded but does not limit their constructor arguments. A few of the types in SAFETOIMPORT have constructors that allocate memory proportional to their input builtins.bytes, builtins.list, builtins.range. A 40-byte pick...

Arbitrary file write via tar traversal in mlflow

A vulnerability in MLflow's pyfunc extraction process allows for arbitrary file writes due to improper handling of tar archive entries. Specifically, the use of tarfile.extractall without path validation enables crafted tar.gz files containing .. or absolute paths to escape the intended extractio...

Improper S3 ownership verification in Bedrock AgentCore Starter Toolkit

SummaryAn issue has been identified in the Bedrock AgentCore Starter Toolkit versions prior to v0.1.13 that may allow a remote actor to inject code during the build process, leading to code execution in the AgentCore Runtime. ImpactA remote actor could inject code during the build process, leadin...

Frigte has broken access control viewer user can delete admin and other users account

SummaryUsers with the viewer role can delete admin and other users account. It this leads to denial of service and affects data integrity. DetailsEndpoint DELETE /api/users/admin is enable to anonymous user. PoCI deleted admin user on demo.frigate.video: ImpactIt this leads to denial of service a...

PyMuPDF has a path traversal in _main_.py

A path traversal and arbitrary file write vulnerability exist in the embedded get function in 'main.py' in PyMuPDF version, 1.26.5...

pypdf has inefficient decoding of array-based streams

ImpactAn attacker who uses this vulnerability can craft a PDF which leads to long runtimes and/or large memory usage. This requires accessing an array-based stream with lots of entries. PatchesThis has been fixed in pypdf==6.9.1. WorkaroundsIf you cannot upgrade yet, consider applying the changes...

Denial of service via non-terminating SYLT frame parsing loop in tinytag

Summarytinytag 2.2.0 allows an attacker who can supply MP3 files for parsing to trigger a non-terminating loop while the library parses an ID3v2 SYLT synchronized lyrics frame. In server-side deployments that automatically parse attacker-supplied files, a single 498-byte MP3 can cause the parsing...

Stored XSS in PySpector HTML Report Generation leads to Javascript Code Execution

SummaryPySpector versions = 0.1.6 are affected by a stored Cross-Site Scripting XSS vulnerability in the HTML report generator. When PySpector scans a Python file containing JavaScript payloads i.e. inside a string passed to eval , the flagged code snippet is interpolated into the HTML report...

PySpector has a Plugin Sandbox Bypass leads to Arbitrary Code Execution

SummaryPySpector versions = 0.1.6 are affected by a security validation bypass in the plugin system. The validateplugincode function in pluginsystem.py, performs static AST analysis to block dangerous API calls before a plugin is trusted and executed. However, the internal resolvename helper only...

multipart vulnerable to ReDoS in `parse_options_header()`

SummaryThe parseoptionsheader function in multipart.py uses a regular expression with an ambiguous alternation, which can cause exponential backtracking ReDoS when parsing maliciously crafted HTTP or multipart segment headers. This can be abused for denial of service DoS attacks against web...

pypdf: manipulated stream length values can exhaust RAM

ImpactAn attacker who uses this vulnerability can craft a PDF which leads to large memory usage. This requires parsing a content stream with a rather large /Length value, regardless of the actual data length inside the stream. PatchesThis has been fixed in pypdf==6.8.0. WorkaroundsIf you cannot...

django-unicorn affected by component state manipulation via unvalidated attribute access

SummaryComponent state manipulation is possible in django-unicorn due to missing access control checks during property updates and method calls. An attacker can bypass the intended ispublic protection to modify internal attributes such as templatename or trigger protected methods. Vulnerability...

FastMCP OAuth Proxy token reuse across MCP servers

While testing the OAuth Proxy implementation, it was noticed that the server does not properly respect the resource parameter submitted by the client in the authorization and token request. Instead of issuing the token explicitly for this MCP server, the token is issued for the baseurl passed to...

ha-mcp OAuth 2.1 DCR mode enables network reconnaissance via an error oracle

SummaryThe ha-mcp OAuth consent form beta feature accepts a user-supplied haurl and makes a server-side HTTP request to haurl/api/config with no URL validation. An unauthenticated attacker can submit arbitrary URLs to perform internal network reconnaissance via an error oracle. Two additional cod...

Magic Wormhole: "wormhole receive" allows arbitrary local file overwrite

ImpactWhat kind of vulnerability is it? Who is impacted?Receiving a file wormhole receive from a malicious party could result in overwriting critical local files, including /.ssh/authorizedkeys and .bashrc. This could be used to compromise the receiver's computer.Only the sender of the file the...

SGLangs `replay_request_dump.py` contains an insecure pickle.load() without validation and proper deserialization

SGLangs replayrequestdump.py contains an insecure pickle.load without validation and proper deserialization. An attacker can take advantage of this by providing a malicious .pkl file, which will execute the attackers code on the device running the script...

MLflow has a command injection in mlflow/sagemaker/__init__.py

A command injection vulnerability exists in mlflow/mlflow versions before v3.7.0, specifically in the mlflow/sagemaker/init.py file at lines 161-167. The vulnerability arises from the direct interpolation of user-supplied container image names into shell commands without proper sanitization, whic...

Graphiti vulnerable to Cypher Injection via unsanitized node_labels in search filters

SummaryGraphiti versions before 0.28.2 contained a Cypher injection vulnerability in shared search-filter construction for non-Kuzu backends. Attacker-controlled label values supplied through SearchFilters.nodelabels were concatenated directly into Cypher label expressions without validation.In M...

ha-mcp has XSS via Unescaped HTML in OAuth Consent Form

SummaryThe ha-mcp OAuth consent form renders user-controlled parameters via Python f-strings with no HTML escaping. An attacker who can reach the OAuth endpoint and convince the server operator to follow a crafted authorization URL could execute JavaScript in the operator's browser. This affects...

vLLM has SSRF Protection Bypass

SummaryThe SSRF protection fix for https://github.com/vllm-project/vllm/security/advisories/GHSA-qh4c-xf7m-gxfc can be bypassed in the loadfromurlasync method due to inconsistent URL parsing behavior between the validation layer and the actual HTTP client. Affected Component- File:...