CVE-2026-58371 is a cross-site scripting (XSS) vulnerability in SeaweedFS versions before 4.30. It arises from improper neutralization of the callback query parameter in JSON endpoints, which are served with Content-Type application/javascript without validation or security headers. This allows a third-party web page to load these endpoints cross-origin via a script tag and read sensitive cluster information such as topology, volume server URLs, gRPC ports, file identifiers, and directory listings. The vulnerability affects multiple unauthenticated endpoints accessible by default configurations. The issue is fixed in SeaweedFS version 4.30.
Red Hat JBoss Enterprise Application Platform 7.3.18 includes security fixes addressing multiple vulnerabilities such as denial of service via malicious JSON Web Encryption tokens, HTTP server failures to reject malformed host headers leading to potential cache poisoning and SSRF, denial of service vulnerabilities with temporary files in Apache CXF, and an HTTP/2 DDoS vulnerability in Undertow. This update replaces version 7.3.17 and is rated as having an important security impact by Red Hat.
Malicious npm package published as part of a coordinated DeFi-themed infostealer campaign. `decimal-format-core` uses a dropper technique: a `postinstall` hook executes `scripts/install-check.cjs` at install time, which fetches a second-stage infostealer payload from the C2 domain `logstream-api.online`. The infostealer harvests cryptocurrency wallet vaults (MetaMask, Phantom, Solflare, OKX, Coinbase, TrustWallet, Backpack, TronLink), Chrome/Firefox/Brave cookies and credentials, SSH keys, AWS credentials, `.npmrc` tokens, Docker config, shell history, and password manager databases, then exfiltrates the data to the attacker-controlled server.
---
_-= Per source details. Do not edit below this line.=-_
## Source: amazon-inspector (41dcb1eea736b0aba6c078a55b8b60553925e6981452e5c4f56e57e419801f87)
On npm install, the package's postinstall script (scripts/install-check.cjs) fetches a JSON config from https://logstream-api.online/config/dfc-sync.json, reads a peerBundle URL from it, downloads a tarball to a temp directory, extracts it into a.peer/ directory, runs `npm install` inside the extracted tree, then require()s the extracted peer-math.js module and invokes syncSession(). The fetched payload is not pinned, hashed, or signature-verified, and the source host is fully attacker-controlled and mutable. This executes arbitrary remote code in the installer's context as a default consequence of `npm install`. The package presents itself with description 'Logform-style numeric and text formatting utilities for Node.js loggers' and keywords (logform, logger, format) that target users searching for the legitimate logform logging library, while the README frames the remote fetch-and-exec as a benign 'Enterprise sync / peer bundle' feature; the path runs by default with no opt-in because resolvePeerBundleUrl falls through to the hardcoded homepage URL when env vars are unset.
## Source: ghsa-malware (7418b3fb88f3b25773cddacaa807900d6e5aab486723f17a4bc6306f1aa247fd)
Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be rotated immediately from a different computer. The package should be removed, but as full control of the computer may have been given to an outside entity, there is no guarantee that removing the package will remove all malicious software resulting from installing it.
JavaScript::Minifier::XS versions before 0.16 for Perl leak memory on every call to minify(), allowing unbounded memory growth.
In JsMinify (XS.xs) the cleanup frees only the NodeSet structures and never the per-token contents buffers allocated in JsSetNodeContents; JsDiscardNode unlinks nodes without freeing their contents. Each token's contents buffer is therefore leaked on every call, and the two early returns taken when the node list is empty leak the whole NodeSet.
A long-lived process that minifies repeatedly, such as an asset pipeline or a server-side minifier endpoint, grows in memory without bound until it exhausts available memory and is killed, causing denial of service.
A directory traversal vulnerability exists in better-helperjs static file server versions up to 3.0.5. The vulnerability arises from improper path validation using string prefix checks, allowing attackers to access files in adjacent directories with names sharing the static root directory's prefix. This flaw enables unauthorized reading of sensitive files outside the intended public directory. The issue is fixed in version 3.0.6 by enforcing exact path boundary checks. No known exploits are reported in the wild.
OpenProject is open-source, web-based project management software. Prior to 17.3.3 and 17.4.1, the HTML sanitizer grants <macro> elements unrestricted data-* attributes via :data wildcard. An attacker injects data-controller="poll-for-changes" into a work package description, causing Stimulus.js to mount a controller that fetches an attacker-uploaded attachment and passes it to renderStreamMessage(). This executes arbitrary Turbo Stream actions — including redirect_to — in every victim's authenticated browser session, redirecting them to an attacker-controlled server. This vulnerability is fixed in 17.3.3 and 17.4.1.
Zed Attack Proxy (ZAP) ViewState add-on before version 4 contains an insecure deserialization vulnerability that allows attackers who control a proxied web server to achieve arbitrary code execution by embedding a malicious serialized Java object in the javax.faces.ViewState HTTP response parameter. The JSFViewState.decode() method base64-decodes the ViewState value and passes it directly to ObjectInputStream.readObject() without a deserialization filter, allowlist, or type restriction, causing the malicious object to be deserialized within the ZAP JVM when the Desktop UI renders the ViewState panel.
---
_-= Per source details. Do not edit below this line.=-_
## Source: amazon-inspector (2dcf100385d43e4bb2e48489a99d1fc2d4f919953963fa0339966bd5270f61b3)
All published JavaScript in dist/ (index.js, client.js, server.js, react.js, and sibling chunks) is heavily obfuscated using javascript-obfuscator (declared as a devDependency) with the hex-identifier and shuffled string-array dispatcher style. The package declares no preinstall/install/postinstall lifecycle hooks; no fetch-and-execute against remote hosts, no reads of installer-secret paths (~/.aws, ~/.ssh, ~/.npmrc, browser profiles), no enumeration of process.env, and no hardcoded credentials or attacker-controlled C2 endpoints were observed in the entry points reviewed. The 'ping/GET/id' keyword co-occurrence flagged in dist/server.js fires inside the obfuscated bundle and is not corroborated by any reachable exfiltration path in the files read. Obfuscation of the entire shipped surface makes the code unauditable and is a legitimate transparency concern for an 'ai-node-agent' library — reviewers should weigh maintainer trust before adopting — but on its own does not meet the threshold for a published block.
## Source: ghsa-malware (0730db02e46f4cfb224880f60bcdcdd43ed4d1bc97c68ee404428f7c592445cb)
Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be rotated immediately from a different computer. The package should be removed, but as full control of the computer may have been given to an outside entity, there is no guarantee that removing the package will remove all malicious software resulting from installing it.
The npm package js-price-client-node version 1.0.0 contains malicious code that exfiltrates environment variables from the user's .env file during installation. The package's postinstall script reads and parses the .env file, then sends the data to a concealed remote server using a base58-encoded URL split across two files. The exfiltration is stealthy, with errors suppressed and fake success responses returned to avoid detection. The package falsely claims to fetch cryptocurrency prices but performs no such function, instead harvesting sensitive credentials such as API keys and passwords.
The npm package 'js-client-node' version 1.4.0 contains malicious code that exfiltrates environment variables from the installer's project during installation. It uses a postinstall hook to read the .env file at the project root and sends all environment variables, including sensitive data like API keys and credentials, to a remote server. The package disguises itself as a legitimate crypto price fetcher, with a misleading README and package metadata. This results in silent leakage of sensitive information upon installation.
The npm package set-cookie-ease version 1.1.5 is a malicious typosquatting package masquerading as the legitimate js-cookie library. It contains code that executes arbitrary attacker-controlled JavaScript in Node.js or server-side rendering environments when Cookies.set is called with expires set to 0 or when document is undefined. This leads to remote code execution (RCE) with full host privileges. The package adds unexpected dependencies (axios and request) to support this behavior, despite advertising no dependencies. There is no official patch or remediation available at this time.
A flaw in Node.js Permission API can cause a local server to be started (via a Unix domain socket), even without the `--allow-net` permission.
This vulnerability affects one supported release line: **Node.js 26**.
A flaw in Node.js HTTP/2 client allows a server to send an unlimited number of ORIGIN frames, which could lead to an Out of Memory error on the client.
This vulnerability affects all supported release lines: **Node.js 22**, **Node.js 24**, and **Node.js 26**.
Two data sources (DICOMWebProxy and DICOMJSON) shipped in the default configuration fetch an arbitrary URL parameter without validation. A global authentication service in OHIF automatically injects the authenticated user's OIDC Bearer token into the resulting requests, sending it to the attacker-controlled server. DICOMweb data sources are not impacted.
Bitwarden Server before 2026.5.0 contains a JSON injection vulnerability in IntegrationTemplateProcessor.ReplaceTokens(), which substitutes user-controlled values into event-integration templates without JSON encoding. When an organization has configured an event integration whose template references a user-controlled token (such as #ActingUserName# or #UserName#, populated from a member's display name), an authenticated member can set their display name to JSON metacharacters and inject arbitrary key-value pairs into the rendered payloads delivered to webhook, SIEM, Slack, Teams, or Datadog endpoints, making injected fields indistinguishable from legitimate template output.
---
_-= Per source details. Do not edit below this line.=-_
## Source: amazon-inspector (e43be65a0930b121cf4d610c70b2d07165e4d335dece4344590b471d078fa5b5)
Package self-describes as a 'System binary configuration tool' but ships a covert screen and clipboard surveillance overlay. pointer.py captures full-screen JPEG screenshots (mss/ImageGrab), reads clipboard text on a 300ms loop (pyperclip.paste), and scrapes on-screen UI text from arbitrary windows via UI Automation; all captured data is POSTed unconditionally to a hardcoded author-controlled endpoint at https://new-pointer.vercel.app/api (pointer.py:33 VERCEL_API_URL). The destination is not configurable. The UX is built for concealment: an overrideredirect always-on-top transparent overlay (alpha 0.75, transparent-white) and hidden hotkeys including ctrl+q panic_exit (os._exit), esc stealth_hide, and 1+` stealth_show. The npm wrapper (index.js startApp) additionally performs a silent 'ghost install' of Python 3.12.3 from python.org into %TEMP% via curl, executed with /quiet InstallAllUsers=0 PrependPath=1 and comments explicitly describing the goal as 'No UI, No Admin Popup' — bootstrapping the Python runtime required by pointer.py without user consent. Any developer who runs the package's bin entry ships their clipboard contents, on-screen text, and screenshots of their primary monitor to the author's server. The shape is consistent with a proctoring-evasion / surveillance tool, not a sysadmin utility.
## Source: ghsa-malware (4cdfeaff9ba72cf95d0e701523049f137b60eadc9b54950e7c20ec3c87b59758)
Any computer that has this package installed or running should be considered fully compromised. All secrets and keys stored on that computer should be rotated immediately from a different computer. The package should be removed, but as full control of the computer may have been given to an outside entity, there is no guarantee that removing the package will remove all malicious software resulting from installing it.
Tiptap for PHP before version 2.1.1 contains an input validation vulnerability that allows authenticated attackers to cause a denial of service by submitting Tiptap JSON with the attrs.href field set to an array instead of a string, causing an unhandled TypeError in the Link::isAllowedUri() function when passed to preg_match(). Attackers can persist malformed JSON records that permanently crash the server-side HTML rendering pipeline for all subsequent viewers of that record until the database entry is manually repaired.