xr-specialist
Pass
Audited by Gen Agent Trust Hub on Jul 5, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill ingests data from local reports which could contain adversarial instructions (Indirect Prompt Injection). The risk is low as the skill lacks executable capabilities such as network access, file writing, or command execution, limiting the impact to the generated report text.
- Ingestion points:
SKILL.md(reading triage-scorecard.md and report outputs) - Boundary markers: Absent
- Capability inventory: None (read-only and report generation)
- Sanitization: Absent
- [SAFE]: No malicious patterns, obfuscation, or dangerous capabilities were detected. The skill uses standard installation methods and focuses on reviewing project metadata.
Audit Metadata