Back

CRITICAL

CodeWhale: run_tests Tool Enables RCE via Malicious Repository Without Approval

Published May 28, 2026

Description

CodeWhale is a DeepSeek + MiMo coding agent in terminal. From 0.3.0 to 0.8.23, the run_tests tool executes cargo test in the workspace with ApprovalRequirement::Auto, meaning it runs without any user approval prompt. cargo test compiles and executes arbitrary code: test binaries, build.rs build scripts, and proc macros. While auto-approving test execution is a deliberate design choice, it creates an inconsistency in the security boundary. However, in a malicious repository, test code can execute arbitrary shell commands, exfiltrate credentials, or establish persistence with zero approval. The attack is amplified by AGENTS.md (auto-loaded into the system prompt), which can instruct the model to run tests proactively at session start. This vulnerability is fixed in 0.8.23.

Affected products

Remediation

No remediation recorded yet.

Metrics

Weaknesses (1)

References (5)

Change history (0)

No recorded changes yet.

Sources
CVE.org / MITRE
Status PUBLISHED
Assigner GitHub_M
Published May 28, 2026
Updated Jun 1, 2026
Reserved May 11, 2026
CISA Vulnrichment
Updated Jun 1, 2026
NVD
Status Deferred
Modified Jun 17, 2026
Red Hat
Severity n/a
Public date n/a
GHSA-WX44-2Q6H-J6P8