container-escape-techniques

Fail

Audited by Snyk on Nov 16, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.30). The skill is an openly advertised offensive container-escape playbook aimed at host systems, not an attack on the user the agent serves, but its persistence backdoors and anti-forensic cleanup steps conceal activity from the system owner, so it is rated medium under the recall-first override.

MEDIUM W013: Attempt to modify system services in skill instructions.

  • Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill is an offensive playbook that directs the agent to escape container isolation, mount and chroot into the host filesystem, read /etc/shadow, create privileged containers via the Docker socket, and overwrite host runc, all of which bypass security boundaries and modify host state.

Issues (2)

E004
CRITICAL

Prompt injection detected in skill instructions.

W013
MEDIUM

Attempt to modify system services in skill instructions.

Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 16, 2026, 12:01 PM
Issues
2