Início / CVEs
Base de CVEs
Vulnerabilidades publicadas, com severidade, exploração conhecida e fontes oficiais.
CVE-2026-114654
Média · 6.5
CWE-681 22/10/2026
BSON output corruption via out-of-range binary subtype bypassing UUID and OldBinary guards
A numeric-type conversion flaw in the BSON handling library of the MongoDB C# Driver allows an out-of-range binary subtype value supplied through untrusted textual input to bypass the library's binary-element integrity checks. When such a value is later encoded, it is narrowed to a smaller wire representation, producing binary elements that misdeclare their type or length. An unauthenticated party able to supply text parsed by an affected application can cause it to emit and persist malformed binary data, which subsequent readers either reject with an error or silently interpret as shorter than the data that was validated.
CVE-2026-114653
Média · 6.5
CWE-185 22/10/2026
Silent client-side encryption bypass via unescaped field-name regex in CSFLE schema builder
A client-side encryption schema construction component in the MongoDB C# Driver places caller-supplied field identifiers into a position that is interpreted as a pattern expression, without normalizing or validating them. As a result, an application that designates certain fields for client-side field-level encryption may have the protection silently fail to apply, so data intended to be encrypted is stored and returned in cleartext; conversely, the rule may apply more broadly than intended. No user action is required to cause the condition, and any party with ordinary read access to the stored data — without possession of the encryption keys — can observe the unprotected values.
CVE-2026-114652
Média · 6.5
CWE-943 22/10/2026
Update operator injection via single-element list rewriting in the update builder
A data-query-logic neutralization weakness in the update-statement construction component of the MongoDB C# Driver can cause application-supplied values to be interpreted as query/update logic rather than as literal data. An application user who can influence the content of values that the application passes into the affected builder surface, without any special privilege beyond ordinary read-write application access, may cause stored data to be modified or removed in ways the application never intended. The result is unauthorized integrity impact on the affected application data; there is no direct loss of confidentiality or availability of the database service.
CVE-2026-114651
Média · 4.0
CWE-908 22/10/2026
Cross-document data leak via malformed BSON array parsing in reused thread buffer
A memory-handling weakness in the BSON deserialization component of the MongoDB C# Driver may cause data left over from previously processed content to be returned when specially crafted serialized input is parsed. An unauthenticated party able to supply serialized data to an application that maps certain primitive collection types can cause residual bytes from unrelated data processed earlier on the same execution thread to be surfaced as legitimate values. This can result in disclosure of a limited amount of information belonging to other requests or principals of the consuming application. No authentication or special privileges within the application are required.
CVE-2026-114649
Média · 6.8
CWE-1025 22/10/2026
Authentication bypass via issuer-and-serial-only certificate equality in shared cluster cache
The MongoDB C# Driver compares TLS client certificates by issuer and serial number only when deciding whether two clients can share cached connections. In an application that creates clients from certificates supplied by different users, a user who provides a certificate copying another user's issuer and serial number, without that user's private key, can be given connections authenticated with the other user's certificate. Database operations can then run under the other user's identity, affecting the confidentiality and integrity of data that identity can access.
CVE-2026-114650
Média · 5.9
CWE-838 22/10/2026
Plaintext exposure of encrypted fields via ANSI code-page corruption of database name
A flaw in the client-side encryption support of the MongoDB .NET/C# driver can cause data designated for client-side field-level encryption to be written without encryption. Under certain platform text-encoding configurations, an internal name conversion loses information, so the encryption layer fails to associate the operation with its configured encryption rules and the operation proceeds without an error being raised. An application user with the ability to write data through an affected application can therefore have sensitive fields persisted in plaintext, defeating the confidentiality guarantee that client-side encryption is intended to provide. No integrity or availability impact was observed.
CVE-2026-114648
Alta · 7.6
CWE-943 22/10/2026
NoSQL injection via dot and dollar characters in LINQ filter field names and dictionary keys
The MongoDB C# Driver does not neutralize dot and dollar characters in dictionary keys, field names and array-filter identifiers used in LINQ queries before placing them into the field paths it sends to the server. An application user whose input reaches these positions can redirect a query or update to fields the application did not name, which can expose other users' documents and widen updates beyond the intended data. Certain values can also write malformed fields that leave stored documents unreadable by the application until they are repaired.
CVE-2026-114647
Média · 4.3
CWE-943 22/10/2026
NoSQL operator injection via unvalidated replacement document in findOneAndReplace
The MongoDB C# Driver does not reject replacement documents whose top-level field names begin with a dollar sign in FindOneAndReplace, unlike other replace operations. An application that builds such a document from user-supplied field names can have the server apply update operators instead of a full replacement. The effect is limited to the document matched by the application's own filter, whose contents can be modified in ways the application did not intend.
CVE-2026-114646
Média · 6.5
CWE-130 22/10/2026
NoSQL command injection via mismatched raw BSON length prefix written verbatim to the wire
A data-validation flaw exists in the BSON handling and binary serialization path of the MongoDB C# Driver. When an application constructs a raw or lazy BSON document object from externally supplied binary data, the driver does not verify that the size quantity declared inside that data matches the amount of data actually provided, and it copies the data onto the wire unchanged. An unauthenticated party who can supply binary content to such an application can thereby cause additional elements to be inserted into, or existing elements to be removed from, the database command the application issues, without needing any database credentials of their own. This can result in unauthorized modification of stored data and disclosure of data outside the intended scope of the application's request, bounded by the database privileges the application itself holds.
CVE-2026-114645
Média · 5.2
CWE-190 22/10/2026
Heap overflow via 32-bit length wrap of the URI username in Windows Kerberos auth message wrapping
A 32-bit integer wraparound in the MongoDB C Driver's Windows-only Kerberos/SSPI authentication code can cause an undersized heap buffer to be allocated when an excessively long credential value is supplied in the client's connection configuration. A party able to control that client-side configuration, in conjunction with the affected authentication mechanism, can cause the driver to write outside the bounds of the allocated buffer. This may result in corruption of client process memory and denial of service of the client application. The issue affects only client builds compiled with the Windows-native authentication backend.
CVE-2026-114644
Média · 5.9
CWE-770 22/10/2026
Memory-exhaustion denial of service via unbounded symbol interning during BSON/JSON deserialization
The MongoDB C# Driver's BSON library contains an uncontrolled resource consumption weakness in its deserialization handling of one BSON value type. An application that deserializes untrusted content with the library retains data derived from that content for the lifetime of the process, with no limit on how much accumulates. An unauthenticated remote party who can cause such content to be deserialized can drive unbounded memory growth, exhausting the host application's available memory and denying service. There is no impact to confidentiality or integrity, and no MongoDB server component is affected.
CVE-2026-114643
Média · 5.4
CWE-694 22/10/2026
Duplicate geometry fields via unvalidated GeoJSON extra members
The MongoDB C# Driver writes GeoJSON extra members without checking whether their names collide with reserved GeoJSON members such as type and coordinates. An application that places user-supplied field names into these extra members can store documents with duplicate geometry fields, so the server and the driver resolve different values for the same record. Affected records can also become unreadable through the driver, interrupting application code that reads them.
CVE-2026-114613
Alta · 7.5
CWE-401 22/10/2026
MessagePack: Unbounded memory leak in msgpack C extension on strict_map_key validation failure
MessagePack is the serializer implementation for Python msgpack.org. Prior to 1.2.3, the msgpack-python C extension (msgpack._cmsgpack) leaks a decoded map value when deserialization rejects a key with strict_map_key=True (the default). Repeated malformed inputs can cause unbounded process memory growth and denial of service in applications that deserialize attacker-controlled MessagePack data. Both unpackb() and the streaming Unpacker API are affected. The pure-Python fallback does not exhibit this leak in the reported tests. This issue is fixed in version 1.2.3.
CVE-2026-114612
Crítica · 9.9
CWE-269 22/10/2026
LiteLLM: Privilege Escalation to Proxy Admin via Cross-Domain Reuse of the Salt Key
LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From 1.87.0 until 1.100.4, 1.101.3, 1.102.2, 1.103.1, and 1.104.0-rc.2, the proxy reuses one cryptographic context for at-rest secrets and session tokens, allowing an authenticated internal user to have crafted metadata containing a forged administrator credential encrypted and returned by the API-key flow. Replaying that ciphertext as a bearer token causes the proxy to trust the forged proxy administrator identity, permitting administrative access and arbitrary commands through the MCP stdio endpoint. Versions 1.87.0 through 1.90.x require EXPERIMENTAL_UI_LOGIN to be enabled, while versions after 1.91.0 are exploitable in the default configuration. This issue was fixed in LiteLLM 1.100.4, 1.101.3, 1.102.2, 1.103.1, and 1.104.0-rc.2.
CVE-2026-114519
Alta · 8.6
CWE-862 22/10/2026
Missing authorization in the Prospero Flow CRM user API allows a basic role user to read and delete other accounts
Missing Authorization (CWE-862) in the user API of Roskus Prospero Flow CRM from 2.0.0 before 5.26.0 allows a user with a basic role to read the accounts of other users of its own company, including accounts with higher privileges, and, in versions from 3.5.0, to delete them, because the routes GET /api/user/{id} and DELETE /api/user/{id} in routes/api.php carry only the auth:api middleware, which checks that a JSON Web Token is valid and nothing more, and neither UserReadController::read() nor UserDeleteController::delete() checks a permission, a policy or a role before acting. Object access is scoped by company: the read path limits the query to the company of the authenticated user, and the delete path compares the company of the target account with the company of the authenticated user. The delete path skips that comparison when the company of the authenticated user is Company::DEFAULT_COMPANY, so a user with a basic role in that company can delete any account of any company of the installation. Deleting an account removes the data held on it, and the last administrative account of a company can be deleted. This gives the attacker the personal data held on the other accounts, including email address, telephone number, postal address and IBAN, and the ability to delete accounts it has no permission to administer.
CVE-2026-113876
Média · 4.3
CWE-863 22/10/2026
SmileCDR HAPI FHIR unauthorized read
SmileCDR HAPI FHIR does not correctly check authorization for conditional create actions. A user with CREATE but not READ permissions could submit a conditional create whose predicate matches an existing resource and receive the target resource's contents.
CVE-2026-113875
Baixa · 3.1
CWE-862 22/10/2026
SmileCDR HAPI FHIR unauthorized read via subscription
On a SmileCDR HAPI FHIR server with REST-hook subscriptions enabled, if a user is granted subscription-creation permissions, they may be able to query and read resources that they are not otherwise permitted to access.
CVE-2025-69723
Média · 6.1
CWE-79 22/10/2026
A reflected cross-site scripting (XSS) vulnerability in localstack v4.9.2 allows attackers to...
A reflected cross-site scripting (XSS) vulnerability in localstack v4.9.2 allows attackers to execute arbitrary Javascript in the context of the user's browser via a crafted URL.
CVE-2025-65200
Crítica · 9.6
CWE-79 22/10/2026
An issue was discovered in localstack 4.9.2 allowing attackers to execute arbitrary code via...
An issue was discovered in localstack 4.9.2 allowing attackers to execute arbitrary code via crafted content hosted on a web-accessable URL and when passed said URL in the the templateURL to the _localstack/cloudformation/deploy endpoint, the hosted content can get executed in the context of the docker container.
CVE-2022-51026
Crítica · 9.8
CWE-290 22/10/2026
Bytebase 0.12.0 before 1.0.1 Authentication Bypass via GitLab Self-Hosted Login
Bytebase 0.12.0 before 1.0.1 contains an authentication bypass vulnerability that allows unauthenticated attackers to log in as any user. The flaw stems from trusting client-supplied GitLab provider configuration in the POST /api/auth/login/GITLAB_SELF_HOST endpoint. Attackers can set instanceUrl to a server they control that returns an active user with a chosen email. Bytebase then issues session tokens for that principal, letting attackers impersonate existing users including the Owner or create new members.
CVE-2026-97635
Alta · 8.1
CWE-79 22/10/2026
Agent Zero Web UI XSS
The Agent Zero Web UI before 1.15 does not correctly sanitize user-provided input in chat messages or file uploads. When Agent Zero reproduces content in the chat interface, any injected JavaScript is executed in the context of the victim’s authenticated Agent Zero Web UI session.
CVE-2026-93336
Alta · 7.1
CWE-862 22/10/2026
Missing authorization in the unit resources download route in Juju controllers
Missing authorization in the unit resources download route in Canonical Juju 4.0 through 4.0.15 allows a controller-authenticated principal (a unit agent, or a user with only login permission) to read any application's charm resource in any model, including container-registry credentials, via a crafted request naming an arbitrary unit.
CVE-2026-77157
Média · 6.4
CWE-353 22/10/2026
ast-cli: Auto-update downloads are not verified against a checksum before execution
ast-cli downloads two helper binaries, the Vorpal ASCA real-time engine and the SCA Resolver, from "download.checkmarx.com" as part of its auto-update flow, triggered by the Checkmarx’s "VSCode/JetBrains" extensions and by cx scan "sca-realtime".
Prior to the fix, the downloaded archive was not verified against a checksum before extraction and execution. An attacker able to substitute the served archive could deliver an arbitrary binary that ast-cli would extract and execute the contents of the archive, since no integrity check was performed on the download.
Successful exploitation requires compromising "download.checkmarx.com" and replacing the served archive with one containing an attacker-controlled binary. If the served archive had been replaced that way, any developer machine or CI pipeline invoking ASCA (via Checkmarx's VSCode/JetBrains extensions) or cx scan "sca-realtime" on an affected version would have extracted and executed that binary in place of the intended download.
The download is now verified against a published SHA-256 checksum before extraction.
CVE-2026-77156
Média · 6.4
CWE-22 22/10/2026
ast-cli: Path traversal in tar extraction on Linux/macOS allows writes outside the install directory
Incomplete validation of extracted archive entry paths in ast-cli's tar extraction logic on Linux and macOS (used when installing the Vorpal ASCA real-time engine and the SCA Resolver) would allow a malicious tar archive containing absolute or traversal path entries to write files outside the install directory. Windows builds were not affected, as the Windows extractor performs this path validation.
Exploitation required substituting the archive served at "download.checkmarx.com". An attacker able to compromise that server and replace the archive could write arbitrary, executable files to arbitrary filesystem paths via this path traversal.
Extraction now validates that each entry's destination path resolves within the working directory and rejects absolute paths.
CVE-2026-54532
Média · 6.9
CWE-829 22/10/2026
CloudPirates Open Source Helm Charts: pull_request_target workflow vulnerable to cache poisoning and OIDC token theft
CloudPirates Open Source Helm Charts is a collection of Helm charts. Prior to commit 271d415, pull_request_target workflow is vulnerable to cache poisoning and OIDC token theft. A pull_request_target workflow checks out fork code using github.event.pull_request.head.sha or .head.ref. This gives attacker-controlled code access to base repository secrets and tokens. A pull_request_target workflow writes to a shared Actions cache. A subsequent release or publish workflow restores from the same cache scope, allowing a poisoned cache entry to execute during release. This issue has been patched via commit 271d415.