In TaskingAI v0.3.0 in the DALL-E 3 image generation tool save_url_image function, a path traversal vulnerability allows attackers to write downloaded images to arbitrary locations on the server filesystem by manipulating the project_id parameter.

TransformerOptimus SuperAGI v0.0.14 is vulnerable to Incorrect Access Control in the tool controller. In affected source snapshots, get_tool and update_tool in superagi/controllers/tool.py accept a caller-supplied tool_id and fail to verify organization ownership through the associated toolkit. A remote authenticated attacker from one organization can read or modify another organization's tool metadata through /tools/get/{tool_id} and /tools/update/{tool_id}.

DeepTutor 1.4.0 Unrestricted Shell Command Injection via WebSocket

deeptutor 1.4.0 contains code injection in ExecTool.execute. Through the live tutorbot WebSocket interface, a remote caller can induce the tool layer to execute reviewer-chosen shell commands in the service environment.

GetSimple CMS: Missing CSRF protection in UpdateCE allows forging a privileged server-side update request

GetSimple CMS is a content management system (CMS), and GetSimple CMS CE is the community edition of that CMS. Prior to version 1.5, the UpdateCE update form contained no anti-CSRF token, and the POST handler performed no token or request-origin verification. A remote attacker can host a page that auto-submits a forged POST to the update endpoint; when an authenticated administrator visits it, the server performs an attacker-directed download-and-deploy operation in the administrator's session — with no further interaction. Because the deployed content is executed (see the related ZIP-extraction advisory), this yields remote code execution. The url field is additionally written into the form unescaped, providing a secondary HTML-injection sink via a malicious upgrade.json. This issue has been patched in version 1.5.

Vvveb: Server-side request forgery in Vvveb via IPv6 bypass of validateUrl() in editor oEmbed proxy

Vvveb is a powerful and easy to use CMS with page builder to build websites, blogs or ecommerce stores. Prior to version 1.0.8.6, Vvveb's SSRF guard resolves a host with an IPv4-only function and never inspects IPv6, so any host that lacks an A record passes a private-range check. Editor oEmbed proxy fetches an attacker-supplied URL server side and reflects a response body, so an authenticated admin-panel user (default role site_admin or higher) can read internal-only services and cloud metadata, including IAM credentials, using an IPv6 literal or a domain that carries only an AAAA record. This issue has been patched in version 1.0.8.6.

Sakai Conversations has a Stored XSS Issue

Sakai is a Collaboration and Learning Environment (CLE). From versions 23.0 to before 23.5, and versions 25.0 to before 25.3, the Sakai Conversations tool stores topic and post messages without HTML sanitization, and the frontend renders them using LitElement's unsafeHTML() directive, resulting in stored cross-site scripting (XSS). Any authenticated user with access to a site that has the Conversations tool enabled can inject arbitrary HTML and JavaScript that executes in the browsers of all other users who view that topic or post. This issue has been patched in versions 23.5, 25.3, and 26.0.

Zebra before 4.4.0 Consensus Split via SIGHASH_SINGLE Missing-Output Handling

Zebra before 4.4.0 contains a consensus divergence vulnerability in V5 transparent signature verification, computing a ZIP-244 digest for SIGHASH_SINGLE inputs lacking corresponding outputs instead of failing. Attackers can craft V5 transactions with fewer outputs than inputs that Zebra accepts and templates via getblocktemplate, producing blocks zcashd rejects.

Zebra before 6.3.0 False Readiness via Discarded One-Hash FindBlocks Response

Zebra before 6.3.0 contains an improper exceptional condition check in ChainSync::obtain_tips that discards valid one-hash FindBlocks responses, falsely reporting close-to-tip status. Peers returning only the next block hash cause a zero-length sync sample, making the /ready endpoint return 200 OK while the node remains behind the tip.

Zebra before 11.0.0 Denial of Service via getblock Verbosity 2

The getblock RPC method in zebra-rpc before 11.0.0, used by the Zcash Foundation's Zebra node, panics on verbosity 2 for a side-chain block because the block's -1 confirmations sentinel is converted to u32 with .expect(), aborting the process. Remote unauthenticated attackers, directly or through lightwalletd, can repeat this call to keep the node in a crash loop.

Zebra before 6.1.0 Incorrect Block Size Calculation in getblocktemplate

Zebra before 6.1.0 contains an incorrect calculation vulnerability in its ZIP-317 block template selector that omits header and transaction-count size from the block budget. Attackers can place valid selectable transactions in a victim miner's mempool to shape templates into oversized blocks, causing rejection and wasted proof-of-work.

Zebra before 6.3.0 Peer Misbehavior Ban Bypass via Gossiped Blocks

Zebra before 6.3.0 contains a protection mechanism failure that allows unauthenticated peers to evade misbehavior scoring by supplying invalid gossiped blocks. The inbound cleanup step wrongly downcasts RouterError to VerifyBlockError and discards the score, so attackers can repeatedly force block download and Equihash verification without being banned.

Ghost 4.39.0 before 6.64.0 Invite Token Disclosure via Admin API

Ghost from 4.39.0 before 6.64.0 contains an information disclosure vulnerability in the Admin API that allows staff users to view secret tokens of pending staff invites. Staff users with invite viewing permission can accept pending invites for higher-privileged roles to escalate their privileges.

Ghost 0.5.0 before 6.64.0 Privilege Escalation via Staff Role Assignment

Ghost 0.5.0 before 6.64.0 does not correctly restrict staff role assignment, allowing users with the Editor or Super Editor role to assign their own role to other staff despite lacking permission to do so. An authenticated Editor or Super Editor can promote Author and Contributor users to Editor or Super Editor.

Filament: Multi-factor authentication (app) management actions do not require password reauthentication

Filament is a collection of full-stack components for accelerated Laravel development. From 4.0.0 until 4.13.3 and 5.8.3, app-based multi-factor authentication management actions do not consistently require confirmation of the current password. An attacker with access to an authenticated user session can set up app-based MFA and obtain recovery codes, or disable app-based MFA and regenerate recovery codes by supplying an existing app code or recovery code, without knowing the account password. Email-based MFA is not affected, and the issue does not independently permit an unauthenticated sign-in, but changing the app-MFA configuration may lock the legitimate user out. This issue is fixed in versions 4.13.3 and 5.8.3.

modelcontextprotocol mcp-server-fetch/mcp-server-everything Fetch Tool server.py fetch_url server-side request forgery

A security vulnerability has been detected in modelcontextprotocol mcp-server-fetch and mcp-server-everything up to 2026.6.4. Affected is the function fetch_url of the file mcp_server_fetch/server.py of the component Fetch Tool. The manipulation of the argument url/path leads to server-side request forgery. The attack may be initiated remotely. The exploit has been disclosed publicly and may be used. The pull request to fix this issue awaits acceptance.

itsourcecode Pet Shop Management System admin_reject_completed.php sql injection

A vulnerability was determined in itsourcecode Pet Shop Management System 1.0. The affected element is an unknown function of the file admin_reject_completed.php. This manipulation of the argument ID causes sql injection. It is possible to initiate the attack remotely. The exploit has been publicly disclosed and may be utilized.

PictShare < 3.7.1 Sensitive Information Disclosure via info API

PictShare before 3.7.1 contains an information disclosure vulnerability that allows unauthenticated attackers to obtain the secret delete_code and uploader metadata by calling the API::info() endpoint which returns the complete raw metadata object without a field whitelist. Attackers can use the publicly visible file hash to retrieve the delete_code via the info API and then invoke the delete API to permanently delete arbitrary files, while also exposing uploader IP, User Agent, remote port, and SHA-1 hash, resulting in loss of content integrity, availability, and uploader privacy.

SiYuan before v3.8.5 Missing Authorization in Save-Path Resolver Endpoints

SiYuan before v3.8.5 contains a missing authorization vulnerability in the getRefCreateSavePath, getShorthandSavePath, and getDocCreateSavePath endpoints that allows read-only publish visitors to learn unpublished notebook box IDs. Attackers with read-only or anonymous publish access can POST any open notebook ID to receive the global save-box ID and save-path template, revealing a hidden notebook's existence and creation time.

Mooncake transfer engine through 0.3.13.post1 Memory Exhaustion via Unbounded Notify Queue

Mooncake transfer engine through 0.3.13.post1 contains a memory exhaustion vulnerability in TransferMetadata::receivePeerNotify that allows unauthenticated attackers to grow process memory without limit. Attackers can repeatedly send notify frames up to 1 MB to the handshake RPC port, filling the uncapped notifys vector until the out-of-memory killer terminates the engine.

djehuty: Authenticated SPARQL injection in session editing allows writing arbitrary RDF triples

djehuty is a research data repository system developed by 4TU.ResearchData. Prior to version 26.3.2, an authenticated depositor can inject arbitrary SPARQL into a state-modifying (DELETE/INSERT) query by supplying a crafted session name, letting them write (and delete) arbitrary triples anywhere in the RDF store. Because the RDF store is shared across all accounts and datasets, this is an integrity compromise of the whole repository's metadata, not just the attacker's own records. Having a logged-in account is a precondition. djehuty allows self-registration via ORCID/SAML, so this is a low barrier in typical deployments. This issue has been patched in version 26.3.2.