CVE-2026-19200: 
Velociraptor vulnerability analysis and mitigation

CVE-2026-19200 is an authorization bypass vulnerability in Rapid7's Velociraptor DFIR platform, specifically in the verify() VQL (Velociraptor Query Language) function. Due to an implementation fault, the function uses the global artifact repository instead of a temporary one, allowing authenticated users to overwrite existing artifacts without the permissions normally required to do so. All Velociraptor versions prior to 0.77.2 are affected. The vulnerability was published on August 24, 2026, with a patch available the same day. It carries a CVSS v3.1 base score of 8.9 (High) (GitHub Advisory, Feedly).

The root cause is classified under CWE-94 (Improper Control of Generation of Code / Code Injection) and CWE-862 (Missing Authorization). The verify() VQL function, intended to validate artifact syntax and other properties, incorrectly references the global artifact repository rather than a sandboxed or temporary repository during its execution. This allows any caller with the NOTEBOOK_EDIT permission — a privilege typically granted to analyst roles — to overwrite existing artifacts in the global repository without the elevated permissions that artifact modification normally requires. The fix, merged in pull request #4962, changes the verify() plugin to use a temporary repository instead (GitHub PR #4962, GitHub Advisory).

Successful exploitation allows an authenticated user with only NOTEBOOK_EDIT permission (e.g., an analyst role) to overwrite artifacts in the global Velociraptor artifact repository, which can affect all users and hunts relying on those artifacts. An attacker could tamper with forensic collection artifacts to suppress evidence, introduce malicious collection logic, or disrupt incident response operations across the entire Velociraptor deployment. The changed scope (S:C) in the CVSS score reflects that the impact extends beyond the attacker's own session to the broader artifact repository shared across the organization (GitHub Advisory, Feedly).

There is no public proof-of-concept exploit and no evidence of in-the-wild exploitation as of the disclosure date (Feedly). The EPSS score is approximately 0.228%, placing it in the 13th percentile for exploitation likelihood within 30 days (GitHub Advisory). The vulnerability is not listed in the CISA Known Exploited Vulnerabilities (KEV) catalog. Exploitation requires an authenticated session with NOTEBOOK_EDIT permission and user interaction (UI:R per CVSS), limiting opportunistic exploitation but making insider threat or compromised analyst account scenarios realistic.

1. Gain access: Obtain credentials for a Velociraptor account with at least the NOTEBOOK_EDIT permission (e.g., an analyst role account, whether legitimately held or compromised).
2. Open a Notebook: Navigate to the Velociraptor web UI and open or create a Notebook, which provides access to VQL execution.
3. Craft a malicious artifact: Prepare a modified version of an existing global artifact (e.g., a forensic collection artifact) containing altered or malicious VQL logic — such as suppressing certain file collection, exfiltrating data, or introducing backdoor collection tasks.
4. Invoke verify() with the crafted artifact: Call the verify() VQL function within the Notebook, passing the crafted artifact definition as input. Due to the implementation fault, this call operates against the global artifact repository rather than a temporary one.
5. Artifact overwritten: The global artifact is silently overwritten with the attacker-controlled version, affecting all subsequent hunts, collections, or queries that reference that artifact across the organization (GitHub Advisory, GitHub PR #4962).

• Logs: Velociraptor audit logs showing artifact modification events attributed to accounts with only NOTEBOOK_EDIT permissions (rather than artifact management permissions); unexpected artifact write operations originating from Notebook sessions.
• Artifact Repository: Unexpected changes to global artifact definitions, particularly modifications to commonly used forensic collection artifacts; artifact version history showing edits by analyst-level accounts.
• Behavioral: Hunts or collections producing anomalous results inconsistent with expected artifact behavior; missing or altered forensic evidence in collection outputs that previously functioned correctly.

Rapid7 has released Velociraptor version 0.77.2, which fixes this vulnerability by changing the verify() plugin to use a temporary repository instead of the global one (GitHub PR #4962). Organizations should upgrade to version 0.77.2 or later immediately. As an interim workaround, restrict NOTEBOOK_EDIT permissions to only highly trusted users and audit recent artifact modifications in the global repository to detect any unauthorized changes (Feedly, Velociraptor Advisory).

The vulnerability was noted in automated vulnerability tracking feeds including VulDB and Offseq Radar shortly after disclosure on August 24, 2026 (VulDB). No significant public researcher commentary or media coverage beyond standard vulnerability database entries has been observed at this time.