Jeudi 8 octobre 2026

thephpleague commonmark

league/commonmark from 2.0.0 before 2.10.2 contains a quadratic-time denial of service vulnerability in the GitHub Flavored Markdown Table extension's TableStartParser::tryStart() block-start scan. Unauthenticated attackers can submit a large paragraph of pipe-free lines not starting with letters, forcing repeated full-buffer strpos scans that exhaust PHP worker CPU.

steveukx git-js · Simple-git project simple-git

simple-git, an interface for running git commands in any node.js application, enables applications to execute Git operations from JavaScript. Prior to 4.0.0, the default blockUnsafeOperationsPlugin does not completely reject configuration includes supplied through customArgs to git.clone(). The missing include.path classification permits Git to load an attacker-controlled configuration file, and the initial remediation does not cover includeIf.<condition>.path, allowing the same file-loading primitive through a conditional include. A loaded configuration can set an executable Git option such as core.sshCommand, which Git invokes during the clone operation with the privileges of the Node.js process. Exploitation requires the application to pass attacker-influenced custom arguments and requires an attacker-controlled file that the process can read. This issue is fixed in 4.0.0.

thephpleague commonmark

league/commonmark versions before 2.9.1 contain multiple denial of service vulnerabilities in fenced code block detection, reference link label lookup, and emphasis delimiter processing that perform super-linear work on crafted input. Attackers can submit specially crafted Markdown with long backtick runs, nested brackets, or delimiter sequences to consume disproportionate CPU time and prevent legitimate requests from completing.

thephpleague commonmark

The league/commonmark (thephpleague/commonmark) library in versions >= 1.5.0 and < 2.9.1 contains quadratic parsing complexity in its SmartPunctExtension and AttributesExtension. When either extension is explicitly registered on the Environment (they are not enabled by default and are excluded from the standard CommonMark and GitHub-Flavored Markdown converters), an unauthenticated attacker can submit small, specially crafted Markdown documents — such as text alternating with unpaired quotes, contiguous runs of block-level attribute blocks, or repeated class attributes — to trigger disproportionate CPU consumption and cause a denial of service. Fixed in 2.9.1.

thephpleague commonmark

commonmark versions from 1.5.0 before 2.10.0 contain a denial of service vulnerability in the AttributesExtension when processing distinctly-named attributes. Attackers can submit Markdown with numerous distinct attribute names to cause quadratic-time attribute merging and filtering, consuming disproportionate CPU resources and preventing legitimate requests from completing.

kimai

Kimai before 2.58.0 contains an authentication bypass vulnerability where password reset links remain valid after password changes because the LoginLink signature covers only the user id, not the password hash. Attackers who intercept or cache a password reset link can use it up to 2 additional times within a 1-hour window to log in as the user even after the legitimate user has changed their password.

better-auth scim

better-auth SCIM versions from 1.5.0 before 1.7.0-beta.4 fail to bind non-organization SCIM providers to their creator by default, allowing authenticated users to manage other users' providers. Attackers can regenerate SCIM bearer tokens, invalidate legitimate tokens, and authenticate to SCIM API routes with the attacker-controlled token.

zereight mcp-gitlab

mcp-gitlab contains a path traversal vulnerability in the job_id parameter of build/index.js that allows attackers to redirect GitLab API requests to arbitrary endpoints. Attackers can supply crafted job_id values like ../../../user to escape the intended path prefix and access arbitrary GitLab API resources using the operator's personal access token.

LimeSurvey

LimeSurvey constructs account password-reset links from the client-supplied HTTP Host header without validating it. The optional allowedHosts allowlist that would constrain this is undefined in the default (and documented) configuration, so LSHttpRequest::checkIsAllowedHost() results in no operation. A remote, unauthenticated attacker who submits a forgotten-password request for a known account (requiring only the target's username and email) with a spoofed Host header causes LimeSurvey to email that account a reset link whose hostname is attacker-controlled while embedding the genuine validation_key. When the recipient or an automated inbound mail-security link scanner dereferences the link, the valid reset token is disclosed to the attacker, who replays it against the legitimate host's newPassword endpoint to set a new password and take over the account.

BerriAI litellm · Red Hat Exploit Intelligence · Red Hat Ansible Automation Platform 2 · Red Hat OpenShift AI (RHOAI)

LiteLLM prior to 1.83.10 allows a user to modify their own user_role via the /user/update endpoint. While the endpoint correctly restricts users to updating only their own account, it does not restrict which fields may be changed. A user who can reach this endpoint can set their role to proxy_admin, gaining full administrative access to LiteLLM including all users, teams, keys, models, and prompt history. Users with the org_admin role have legitimate access to this endpoint and can exploit this vulnerability without chaining any additional flaw.

gitoxide · gitoxide gix-submodule · gitoxide gix

gix-submodule before 0.29.0 (gitoxide before 0.5.21, gix before 0.84.0) incorrectly validates the update field in .gitmodules, allowing attackers to bypass the CommandForbiddenInModulesConfiguration guard when a submodule has been initialized with only partial configuration in .git/config. An attacker can inject arbitrary shell commands via the update field in .gitmodules that will be executed when Submodule::update() is called on a previously-initialized submodule, enabling remote code execution.

Gitea

Gitea prior to 1.27.0 contains a server-side request forgery vulnerability that allows authenticated attackers to bypass SSRF protections by exploiting HTTP fetch operations in migration and OAuth avatar code paths that use Go's default http.Get without a custom DialContext. Attackers can supply arbitrary URLs through release asset download URLs, pull-request patch URLs, or OAuth avatar endpoints to reach internal services, cloud instance-metadata endpoints, or read local files such as the application configuration containing database credentials and signing secrets, with exfiltrated content persisted as migration release assets for later retrieval.

budibase server

Budibase versions 3.41.0 before 3.45.0 contain an authentication bypass in the OIDC/SSO login path of @budibase/server. In sso.authenticate, when no existing user matches the incoming SSO subject, the server looks up pending user invites by the IdP-asserted email address alone — without validating an invite code and without an email_verified check (the email_verified gate protects only the existing-account lookup). An attacker who can register at an IdP that the tenant trusts for OIDC and assert a victim's invited email address (even with email_verified=false) claims the pending invite and inherits all of its granted privileges, including builder and admin.global, with no admin exclusion. This results in takeover of the invited principal and, for admin invites, full tenant compromise (access to all apps, datasources including production credentials, and automations); the invite is consumed, denying onboarding to the legitimate invitee.

Red Hat Directory Server 11.5 E4S for RHEL 8 · Red Hat Directory Server 11.7 E4S for RHEL 8 · Red Hat Directory Server 11.9 for RHEL 8 · Red Hat Directory Server 12.2 E4S for RHEL 9

A stack buffer overflow flaw was found in 389 Directory Server (389-ds-base). The get_ruvelement_from_berval() function in repl5_ruv.c copies digit characters from a network-supplied RUV berval into a fixed 16-byte stack buffer without bounds checking. A remote unauthenticated attacker can crash the LDAP server by sending a crafted StartNSDS50ReplicationRequest extended operation containing a replica ID field with more than 16 digit characters. The overflow occurs during payload decoding, before any authorization check. Stack protectors limit impact to denial of service.

JetBrains TeamCity

In JetBrains TeamCity before 2026.2.1 missing validation of Git submodule URLs allowed reading local repositories on the server

OpenSSL · Openssl openssl

Issue summary: A TLS server that calls SSL_set_SSL_CTX() to switch a connection to a different SSL_CTX part way through a handshake may access memory beyond the end of an internal array if the replacement context knows about more provider signature algorithms than the context the connection was created from. Applications which never call SSL_set_SSL_CTX() are not affected. Impact summary: A remote peer may be able to cause a small out-of-bounds read, and in some circumstances a fixed-value out-of-bounds write, on the server heap. This may lead to a Denial of Service. CWE: CWE-787: Out-of-bounds Write Description: A TLS connection records how many certificate slots it has when it is created, taken from the SSL_CTX that created it: the built-in certificate types plus one slot for each provider TLS-SIGALG entry that context was aware of. That count sizes an internal array of per-slot certificate validity flags. An application may replace a connection's SSL_CTX part way through the handshake by calling SSL_set_SSL_CTX(), most commonly from a servername callback in order to serve a different virtual host. Doing so did not refresh the recorded count. A provider signature algorithm's slot index is its position in the list of whichever context resolves it, so if the replacement context is aware of more of them than the original, an algorithm offered by the peer can resolve to an index beyond the end of the array. Processing the peer's signature algorithms then reads one four byte word past the end for each such algorithm and, where the word read is zero, writes a fixed value over it. A peer offering many of them can corrupt heap metadata and abort the process. Only provider signature algorithms which occupy one of the excess slots, and which the server also has configured, have this effect. Codepoints the replacement context does not recognise are discarded without being resolved to a slot, and provider signature algorithms are usable only from TLS 1.3. The two contexts must therefore be aware of different numbers of provider signature algorithms, which requires separate library contexts, a provider loaded between the two being created, or providers which differ in what they advertise - in 4.0, for example, the default provider advertises SM2 where the FIPS provider does not. A deployment meeting the condition is also unable to negotiate the affected algorithms with legitimate clients, since the same stale count hides the corresponding certificates, so the misconfiguration is likely to be noticed. For that reason, and because the configuration is not the default, this issue has been assessed as Low severity. FIPS impact: no No FIPS modules are affected by this issue as the affected code is outside the OpenSSL FIPS module boundary.

OpenSSL · Openssl openssl

Issue summary: QUIC process may keep memory for QUIC packet buffer for much longer period than necessary. Impact summary: Remote peer can exploit this vulnerability by sending maliciously crafted packets, making the local QUIC stack to keep the memory for packet buffers allocated. The time for which the memory remains allocated is entirely under the control of the potentially malicious remote peer. CWE: CWE-770: Allocation of Resources Without Limits or Throttling Description: To save copy operation from the packet buffer to the stream reassemble buffer the QUIC stack leaves the stream data on the packet buffer waiting to be copied to a buffer provided by the local receiving application. The QUIC stack releases a reference to the packet buffer only after the data are copied to the application buffer. This design is more efficient for legitimate data transfers but enables an attacker to allocate a lot more memory than actually required by the data kept in the receiving stream buffer. To mitigate the vulnerability, the QUIC stack now calculates and monitors memory overhead for every stream. The memory overhead for a single stream frame is calculated as a difference between the size of the whole packet that carries the stream frame and the size of the stream frame itself. The memory overhead for a single stream frame is added to the total (cumulative) memory overhead QUIC stack keeps for each stream. Once the cumulative memory overhead exceeds 64kB, the QUIC stack moves the stream frame data from the packet buffer to the stream buffer, starting with the next packet received. FIPS impact: no The FIPS module is not affected as the QUIC implementation is outside of the OpenSSL FIPS module boundary.