xr-cockpit-interaction-specialist
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No security issues detected. The skill serves as a specialized persona for XR cockpit design and interaction. It does not request access to dangerous tools, network resources, or sensitive files.
- [COMMAND_EXECUTION]: The provided shell script
scripts/run.shis a standard wrapper used to invoke the agent runner. It does not perform any suspicious command execution, privilege escalation, or persistence operations. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided tasks via the command line or standard input. While it lacks explicit boundary markers, the absence of high-risk capabilities (such as file system modification or network requests) minimizes the risk associated with this attack vector.
- Ingestion points: User tasks provided via CLI arguments or stdin in
scripts/run.sh. - Boundary markers: Absent; the instructions do not use delimiters to isolate user data from the system prompt.
- Capability inventory: Limited to architectural design and UX guidance; no subprocess calls, file-write, or network operations are present in the skill.
- Sanitization: None identified.
Audit Metadata