Translation in progress Some entries still include text in the original language.
BackSource Hmbown/CodeWhale GHSA-wrj3-vj8c-784f(credit sai-sh);CVE-2026-75858
Hmbown/CodeWhale GHSA-wrj3-vj8c-784f(credit sai-sh);CVE-2026-75858
Indexed by this site 2026-10-05 08:22 8 days ago
Originally published Date not given
Selected 2026-10-05 08:22
Category Event Hmbown/CodeWhale GHSA-wrj3-vj8c-784f(credit sai-sh);CVE-2026-75858 · Hmbown/CodeWhale GHSA-wrj3-vj8c-784f(credit sai-sh);CVE-2026-75858 · Indexed by this site 2026-10-05 08:22 · Originally publishedDate not given SelectedAttention57
CodeWhale 修复 rlm_eval 绕过审批策略的任意代码执行漏洞 rlm_eval auto-approves arbitrary Python execution, bypassing the user's approval policy (RCE)
AI summary
CodeWhale 维护者确认并修复了 rlm_eval 工具的任意 Python 代码执行漏洞,0.8.64 版本在 commit 57f3c89 中完成修补,用户应升级到 0.8.64 或更高版本。该工具的 approval_requirement() 返回 ApprovalRequirement::Auto,引擎将其视为永不询问,从而绕过用户配置的 --approval-policy(on-request、unless-trusted、never),且不产生审批事件。报告指出这与此前已在 run_tests 工具上修补的 CVE-2026-45311 属同一缺陷,修复未覆盖 rlm_eval 和 rlm_open,后两者暴露面更大。
Why it was selected
披露了 CodeWhale 工具审批策略被绕过的完整链路与修复版本,可对照检查自家 Agent 的工具权限设计。
Read original github.com
Why it was selected SelectedAttention57
Attention 0–100: content value contributes up to 70 and circulation up to 30; circulation decays over time. How it was rated
披露了 CodeWhale 工具审批策略被绕过的完整链路与修复版本,可对照检查自家 Agent 的工具权限设计。