A second Claude Code sandbox flaw let blocked network traffic through. Reproduction, affected versions, and the parsing mismatch behind the bypass.
How GitHub comments exposed credentials in Anthropic Claude, Google Gemini, and Microsoft Copilot workflows. Includes reproductions, vendor responses, and the ACM CCS 2026 paper.
TL;DR Agent SkillSlip is a class of path traversal vulnerabilities in AI agent skill/plugin installers. The name field inside skill metadata is used directly in path.join() without validation, writing files to attacker-controlled locations — but the user only sees the archive filename or repository URL, not the internal metadata Found across three tools: Gemini CLI, Claude Code, and Vercel’s add-skill Impact ranges from VS Code terminal hijacking to SSH key injection add-skill fixed in PR #8 and PR #108. Gemini CLI and Claude Code remain unpatched as of writing The Pattern After installing a Gemini CLI skill: .vscode/settings.json injected, terminal hijacked ...
allowedDomains: [], “Empty array = no network access.” — Anthropic Sandbox Runtime Documentation The implementation did not match the documentation. When I configured Claude Code’s sandbox with allowedDomains: [], expecting complete network isolation, the sandbox was wide open and allowed connections to any server on the internet. Anthropic patched this quietly in Claude Code v2.0.55 with a changelog entry saying “Fix proxy DNS resolution” — no mention of a critical security flaw. They assigned CVE-2025-66479 to their runtime library but did not assign a CVE to their flagship product Claude Code. The changelog did not include a security advisory. In practice, the issue was fixed quietly and most users were unlikely to realize there was a security patch. ...