# CVE-2026-110205: Roo Code command approval bypass via glob character masking

Source: https://junglewise.ai/threats/cve-2026-110205-roo-code-through-3-54-0-contains-a-command-approval-bypass  
Publisher: Junglewise Threat Intelligence, https://junglewise.ai/threats  
Published: 2026-10-13. Updated: 2026-10-16  
License: CC BY 4.0 with a link back

CVE-2026-110205 · Severity: high · CVSS 8.8 · Published 2026-10-13

Roo Code is a development assistant that auto-approves shell commands against a security allowlist when configured to do so. An attacker can craft commands with glob characters (*, ?) adjacent to hidden command substitutions to bypass this allowlist check and execute arbitrary commands without user approval. This allows an attacker using prompt injection to execute malicious commands directly on the developer's system.

Technologies: [Roocode Roo Code](https://junglewise.ai/threats/technologies/roocode-roo-code). Vendors: [Roocode](https://junglewise.ai/threats/vendors/roocode).

## Quick answers

**Is CVE-2026-110205 exploited in the wild?** No. Junglewise Threat Intelligence has no report of CVE-2026-110205 being exploited in the wild.

**How severe is CVE-2026-110205?** High severity, with a CVSS score of 8.8, published 13 October 2026.

**Which products and versions are affected?** Roo Code through 3.54.0.

**Is there a fix for CVE-2026-110205?** No fixed version is recorded yet. Check the vendor's advisory in the references below. If you run it, Junglewise drafts a mitigation that fits your setup and applies it once you approve, with senior security engineers behind it.

**How is CVE-2026-110205 attacked?** It is attacked over the network, in a software library.

**Am I affected by CVE-2026-110205?** Check whether you run Roo Code or Roocode Roo Code at an affected version. Junglewise does that check for you around the clock: connect your code, cloud and domains, and it matches new vulnerabilities like this one against what you actually run and tells you when one reaches you.

## Technical details

The parseCommandLine() function in parse-command.ts replaces $(...) command substitutions with placeholders to check them against approval lists, but when a glob character (* or ?) appears adjacent to the placeholder, shell-quote tokenizes it as a glob operator token whose pattern field contains the placeholder. The operator token handler discards the pattern field and only preserves the operator name ("glob"), removing the placeholder from inspection. An attacker can craft commands like echo *$(...) that match allowlisted prefixes but execute denied commands via the discarded substitution when auto-approval is enabled.

## Affected products

- Roo Code Roo Code through 3.54.0

## Timeline

- 2026-10-13: disclosed

## References

- [https://github.com/7rah/oss-cve/blob/6adb836d24ca446fd168c3acdc4c170beb77fe0f/roocode-glob-token.md](https://github.com/7rah/oss-cve/blob/6adb836d24ca446fd168c3acdc4c170beb77fe0f/roocode-glob-token.md)
- [https://github.com/RooCodeInc/Roo-Code/blob/v3.54.0/src/shared/parse-command.ts](https://github.com/RooCodeInc/Roo-Code/blob/v3.54.0/src/shared/parse-command.ts)

## Related threats

- [Roo Code command approval bypass via command substitution](https://junglewise.ai/threats/cve-2026-110206-roo-code-through-3-54-0-contains-a-command-approval-bypass)
- [Roo Code command approval bypass via bash funsub injection](https://junglewise.ai/threats/cve-2026-110204-roo-code-through-3-54-0-contains-a-command-approval-bypass)
- [Roo Code command approval bypass via quote masking](https://junglewise.ai/threats/cve-2026-110203-roo-code-through-3-54-0-contains-a-command-approval-bypass)
- [Roo Code command approval bypass via shell redirect](https://junglewise.ai/threats/cve-2026-110202-roo-code-through-3-54-0-contains-a-command-approval-bypass)
- [Roo Code is an AI-powered autonomous coding agent that lives in users' editors. Prior to version 3.26.7, Due to an…](https://junglewise.ai/threats/cve-2025-65946-roo-code-is-an-ai-powered-autonomous-coding-agent-that-lives-in)