CVE-2026-75858
Received Received - Intake

Remote Code Execution in CodeWhale

Publication date: 2026-08-18

Last updated on: 2026-08-18

Assigner: VulnCheck

Description
CodeWhale (packages codewhale / codewhale-tui) versions >= 0.8.41 and < 0.8.64 contain a remote code execution vulnerability in the rlm_eval tool. The tool's approval_requirement() returns ApprovalRequirement::Auto, which the engine treats as 'never prompt,' causing arbitrary model-supplied Python code to run in a python3 interpreter without consulting the user's configured --approval-policy and without any approval prompt or audit step. An attacker can induce the agent to execute arbitrary code via prompt injection in untrusted content the agent reads (a web page, fetched URL, repository file, or MCP tool result); the companion rlm_open tool can stage such content. Code runs on the user's machine at the user's privilege level. Fixed in 0.8.64.

CVSS Scores

EPSS Scores

Probability:
Percentile:

Meta Information

Published
2026-08-18
Last Modified
2026-08-18
Generated
2026-09-07
AI Q&A
2026-08-18
EPSS Evaluated
2026-09-06
NVD
CVE-2026-75858
EUVD
EUVD-2026-60965

Affected Vendors & Products

Showing 1 associated CPE
Vendor Product Version / Range
codewhale codewhale-tui From 0.8.41 (inc) to 0.8.64 (exc)

Helpful Resources

Exploitability

CWE
CWE Icon
KEV
KEV Icon
CWE ID Description
CWE-94 The product constructs all or part of a code segment using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the syntax or behavior of the intended code segment.

Attack-Flow Graph

AI Quick Actions

Instant insights powered by AI
Executive Summary

This vulnerability is a remote code execution flaw in CodeWhale's rlm_eval tool affecting versions 0.8.41 to 0.8.63. The tool's approval_requirement() function returns ApprovalRequirement::Auto, which the engine interprets as 'never prompt,' allowing arbitrary Python code from models to run without user approval or audit. Attackers can exploit this via prompt injection in untrusted content like web pages or files.

Detection Guidance

Check CodeWhale version with 'npm list codewhale' or 'codewhale --version'. If version is between 0.8.41 and 0.8.63, the system is vulnerable. Inspect logs for unexpected Python interpreter executions or rlm_eval tool usage.

Impact Analysis

This vulnerability allows attackers to execute arbitrary code on your machine at your privilege level. This could lead to unauthorized data access, file manipulation, network exploitation, or installation of malware. Since the code runs without user prompts, you may not even realize it's happening.

Compliance Impact

This vulnerability could lead to unauthorized code execution on user systems, potentially enabling data exfiltration, manipulation, or destruction. For GDPR, this may violate principles of data protection by design and default, risking unauthorized access to personal data. Under HIPAA, it could compromise protected health information integrity and confidentiality if exploited on systems handling such data.

Mitigation Strategies

Upgrade CodeWhale to version 0.8.64 or later immediately. Disable or remove the rlm_eval tool if not needed. Review and restrict access to untrusted content sources like web pages or files processed by the tool.

Chat Assistant

Ask questions about this CVE
Hi! I’m here to help you understand CVE-2026-75858. Ask me anything about the vulnerability, its impact, or mitigation strategies.
0/70

EPSS Chart