Loading vulnerability…
CVE-2026-67340 is a remote code execution vulnerability in ArcadeDB arcadedb-engine before version 26.7.2. The flaw is caused by ScriptTriggerExecutor permitting trigger scripts to resolve host classes from the java.lang package through Java.type. Because java.lang.* is included in the allowed packages, an authenticated user with UPDATE_SCHEMA permission can define a malicious JavaScript trigger that reaches dangerous Java runtime functionality such as Runtime or ProcessBuilder. When the trigger executes, the attacker can cause operating system commands to run on the host where ArcadeDB is deployed.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos.
This repository is a small standalone exploit for CVE-2026-67340 affecting ArcadeDB before 26.7.2. It contains one Python exploit script and a README describing the vulnerability, usage, and remediation. The exploit is post-authentication and abuses ArcadeDB's HTTP API to create a JavaScript trigger whose GraalVM host access permits Java.type("java.lang.Runtime") and therefore Runtime.exec(). Repository structure is minimal: README.md documents the bug and exploit flow; exploit.py is the only code file and the clear entry point. The script uses only Python standard library modules (argparse, base64, json, time, urllib) and performs three authenticated POST requests to /api/v1/command/<db>: create a document type, create a malicious BEFORE CREATE trigger bound to that type, and insert a record to trigger execution. Authentication is Basic auth, with defaults of root / playwithdata and database mydb. Main exploit capability is arbitrary OS command execution as the ArcadeDB process user. The payload is operational rather than just demonstrative because it accepts attacker-supplied commands and can generate a bash reverse shell to an arbitrary HOST:PORT. The command is base64-wrapped before embedding into the JavaScript/SQL payload to avoid quoting issues across JSON, SQL, and JavaScript layers. The exploit does not wait for process completion, so the HTTP request returns quickly while the spawned child process continues running. Fingerprintable targets and behaviors include the ArcadeDB HTTP API on port 2480, the /api/v1/command/<db> execution endpoint, the /api/v1/databases identification endpoint mentioned in the README, use of /bin/bash on the target, and outbound reverse-shell connectivity via bash /dev/tcp. Overall, this is a real authenticated RCE exploit, not a detector or fake sample.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
9 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.