learn-docker-k8s-game

Warn

Audited by Gen Agent Trust Hub on Sep 12, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONPRIVILEGE_ESCALATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill clones a repository from a third-party GitHub user: https://github.com/ericboy0224/learn-docker-and-k8s.git. It also fetches the kind utility from a well-known service at kind.sigs.k8s.io.
  • [REMOTE_CODE_EXECUTION]: The AI is instructed to execute multiple shell scripts from the downloaded repository, including environment checks (engine/environment-check.sh), cleanup routines (engine/cleanup.sh), and various verification scripts (curriculum/**/verify.sh).
  • [COMMAND_EXECUTION]: The skill performs extensive system operations using docker, kubectl, and curl to manage containers, networks, and Kubernetes clusters.
  • [PRIVILEGE_ESCALATION]: Troubleshooting instructions recommend commands requiring administrative privileges, such as sudo systemctl start docker and sudo usermod -aG docker $USER.
  • [INDIRECT_PROMPT_INJECTION]: The skill's narrative and operational logic are loaded from external markdown files within the cloned repository, which could be used to influence the AI's behavior. Ingestion points: Reads AGENTS.md and engine/*.md from the cloned repository; Boundary markers: Absent; Capability inventory: Subprocess execution for shell scripts, Docker, and Kubernetes management; Sanitization: Absent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Sep 12, 2026, 11:45 PM
Security Audit — agent-trust-hub — learn-docker-k8s-game