realtime-graphics-prompt
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely focused on text generation and does not perform any network requests, file writes, or system modifications. Analysis of the instructions and the accompanying playbook confirms that all logic is contained within the prompt engineering domain.
- [SAFE]: No credential exposure or hardcoded secrets were found in the skill or the reference documentation.
- [SAFE]: The skill references local files (e.g.,
references/playbook.md) for context, which is standard behavior for knowledge-based agent skills and does not involve external data exfiltration. - [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface involving user-supplied game concepts and fantasies (ingested in
SKILL.md), which are interpolated into a generated markdown block. While this represents a theoretical attack surface, it is considered safe in this context as the skill generates a static text deliverable for the user to review and manually copy into another environment, providing a natural human-in-the-loop safety boundary. Boundary markers (markdown code blocks) are used for the output.
Audit Metadata