roblox-triage
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed to manage development metadata and state machines. No malicious patterns such as remote code execution, base64 obfuscation, or persistence mechanisms were detected.
- [PROMPT_INJECTION]: The skill triages external data (bug reports, slices, GitHub issues), which represents a standard indirect prompt injection surface for this type of agent. The instructions mitigate this by requiring the agent to "Wait for direction" after making recommendations and enforcing playtest reproduction steps in Roblox Studio, ensuring the agent validates findings rather than blindly following input descriptions.
- [DATA_EXFILTRATION]: No unauthorized data access or external transmission patterns were found. The skill references local configuration paths (e.g.,
docs/agents/triage-labels.md) and standard development tools (roblox-studio-workflow), which are appropriate for its stated purpose. - [EXTERNAL_DOWNLOADS]: The skill mentions a setup command
setup-roblox-skills, but this appears to be a local tool initialization rather than an unverified remote download or piped execution from the internet.
Audit Metadata