CVE-2026-25046 | [Kimi VS Code] Command Injection in publish scripts vsix-publish.js and ovsx-publish.js

Kimi Agent SDK is a set of libraries that expose the Kimi Code (Kimi CLI) agent runtime in applications. The vsix-publish.js and ovsx-publish.js scripts pass filenames to execSync() as shell command strings. Prior to version 0.1.6, filenames containing shell metacharacters like $(cmd) could execute arbitrary commands. Note: This vulnerability exists only in the repository's development scripts. The published VSCode extension does not include these files and end users are not affected. This is fixed in version 0.1.6 by replacing execSync with execFileSync using array arguments. As a workaround, ensure .vsix files in the project directory have safe filenames before running publish scripts.

Published: 2026-01-29 Last update: 2026-04-15 Assigner: [email protected] Source: [email protected]

Conclusion & alert: CVE-2026-25046 is rated Low Risk (11.8/100): CVSS Low severity, with low exploitation likelihood (EPSS 0.01%). Mandatory action: Low composite risk—no urgent action required; patch on your normal maintenance cycle and revisit priority if CVSS or EPSS increases.

Risk is dynamic; we continuously reassess and refresh what is shown on this page as upstream context changes.

Exploit prediction scoring system (EPSS) score for CVE-2026-25046

EPSS lead: Daily EPSS estimates relative likelihood of exploitation; percentile ranks this CVE among scored vulnerabilities (higher = more severe relative rank).

# Date Old EPSS score New EPSS score Delta (New - Old)
1 2026-01-30 0.01%

Full EPSS history (1 record total)

Common vulnerability scoring system (CVSS) metrics for CVE-2026-25046

CVSS metrics for this CVE.

Base score Version Severity Vector Exploitability Impact Score source
2.9 3.1 LOW
CVSS:3.1/AV:L/AC:H/PR:H/UI:R/S:U/C:L/I:L/A:N Click to expand
Attack vector (AV:L)
They already need access on the box, or another person has to do something wrong; it’s not a remote drive-by.
Attack complexity (AC:H)
Even with access, the exploit needs extra luck, timing, or a fussy environment to actually work.
Privileges required (PR:H)
They need powerful rights—admin, root, or similar—before this pays off.
User interaction (UI:R)
A real person has to do something—click, install, enable—otherwise it doesn’t land.
Scope (S:U)
Damage stays in the same “trust bubble” as the broken component—no big spill into unrelated systems.
Confidentiality (C:L)
Some sensitive info could get out, but not a total data dump.
Integrity (I:L)
Attackers could change some data, but it’s limited—not everything goes.
Availability (A:N)
Service keeps running; no real outage angle.
0.3 2.5 [email protected]

Weakness enumeration for CVE-2026-25046

Affected software / configurations for CVE-2026-25046

Vendor Product Version Raw CPE
No affected products in dataset.

References for CVE-2026-25046

cvelogic Threat Intelligence