Remote code execution via unsafe matrix eval
Published Apr 23, 2026 · Updated Apr 23, 2026
Eval injection in ByteDance verl 0.1 through 0.7.0 allows remote attackers to execute code through poisoned model output. In prime_math/grader.py, math_equal calls Python eval twice on an untrusted matrix-form prediction instead of parsing it as data. Exploitation requires prime_math scoring, a matrix reference, and specially formatted model output; successful code runs with the training process's privileges.
Summary
What happened
Eval injection in ByteDance verl 0.1 through 0.7.0 allows remote attackers to execute code through poisoned model output. In prime_math/grader.py, math_equal calls Python eval twice on an untrusted matrix-form prediction instead of parsing it as data. Exploitation requires prime_math scoring, a matrix reference, and specially formatted model output; successful code runs with the training process's privileges.
The record
- CVE
- CVE-2026-6878
- Published
- Apr 23, 2026
- Updated
- Apr 23, 2026
- Vendor
- ByteDance
- Product
- verl
- Classifications
- CWE-265, CWE-264, T1059.006
- Attack vector
- network
- Privileges
- unauthenticated
Timeline
How it unfolded
- Apr 23, 2026CVE publishedPublication date reported by the CVE source.
- Apr 23, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=0.1
- Affected versionversion=0.2
- Affected versionversion=0.3
- Affected versionversion=0.4
- Affected versionversion=0.5
- Affected versionversion=0.6
- Affected versionversion=0.7.0
What conditions does exploitation require?
What is affected?
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- verl math_equal unsafe eval proof of conceptproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo