oma-image

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs agents to run a gh api call to fetch README content from external GitHub repos (e.g. https://github.com/YouMind-OpenLab/awesome-nano-banana-pro-prompts and the gh api example: gh api repos/YouMind-OpenLab/awesome-nano-banana-pro-prompts/contents/README_ko-KR.md) at runtime and to use that fetched content to shape/amplify prompts, so external content directly controls agent prompts.

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 explicitly instructs using a provider CLI flag that bypasses permissions ("agy --dangerously-skip-permissions") and mandates auto-forwarding absolute filesystem paths (automatic --reference forwarding and writing to arbitrary output dirs), which encourages permission bypass and uncontrolled file access/writes on the host.