CVE-2026-75858
Remote code execution via rlm_eval auto‑approval bypass in CodeWhale 0.8.41‑0.8.63.
- CVSS 8.5
- CWE-94
- Design Defects
- Remote
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.
- CWE
- CWE-94
- CVSS base score
- 8.5
- Published
- 2026-08-18
- OWASP
- A03 Injection
- Orthogonal defect classification
- Checking
- Code defect classification
- Missing Check
- Category
- Design Defects
- Subcategory
- Poorly Designed Access Controls
- Accessibility scope
- Remote
- Impact
- Arbitrary Code Execution
- Affected component
- codewhale
- Fixed by upgrading
- Yes
Solution
Upgrade CodeWhale to version 0.8.64 or later.
Vulnerable code sample
from codewhale.rlm_eval import Engine, ApprovalRequirement
def process_model_output(model_code: str) -> None:
# VULNERABLE: ignores user approval policy and executes arbitrary code
engine = Engine()
# The engine treats Auto as never prompt, so any code is run
engine.approval_requirement = lambda: ApprovalRequirement.Auto
engine.run_python(model_code) # executes model_code without promptPatched code sample
from codewhale.rlm_eval import Engine, ApprovalRequirement
def process_model_output(model_code: str) -> None:
# FIX: enforce explicit approval before execution
engine = Engine()
# Use the user's configured approval policy instead of Auto
engine.approval_requirement = lambda: ApprovalRequirement.UserPrompt
# The engine will now require user approval before running code
engine.run_python(model_code)Cite this entry
@misc{vaitp:cve202675858,
title = {{Remote code execution via rlm_eval auto‑approval bypass in CodeWhale 0.8.41‑0.8.63.}},
author = {Bogaerts, Fr\'ed\'eric and Ivaki, Naghmeh and Fonseca, Jos\'e},
year = {2026},
note = {VAITP Python Vulnerability Dataset, entry CVE-2026-75858},
howpublished = {\url{https://netpack.pt/vaitp/vulnerability/CVE-2026-75858/}}
}
Introducing the "VAITP dataset": a specialized repository of Python vulnerabilities and patches, meticulously compiled for the use of the security research community. As Python's prominence grows, understanding and addressing potential security vulnerabilities become crucial. Crafted by and for the cybersecurity community, this dataset offers a valuable resource for researchers, analysts, and developers to analyze and mitigate the security risks associated with Python. Through the comprehensive exploration of vulnerabilities and corresponding patches, the VAITP dataset fosters a safer and more resilient Python ecosystem, encouraging collaborative advancements in programming security.
The supreme art of war is to subdue the enemy without fighting.
Sun Tzu – “The Art of War”
:: Shaping the future through research and ingenuity ::
