kanban-board
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and provided example files do not contain any malicious patterns or security vulnerabilities.
- [PROMPT_INJECTION]: Analysis of the instructions confirms they are restricted to UI layout and data mapping, with no attempts to bypass safety guardrails or override agent behavior.
- [DATA_EXFILTRATION]: No network calls, credential leaks, or sensitive local file path references were detected.
- [REMOTE_CODE_EXECUTION]: The skill produces self-contained HTML and CSS artifacts and does not attempt to download or execute external code or packages.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external design files (DESIGN.md) and user briefs to generate UI components. However, because the skill has no access to sensitive system resources, network capabilities, or file-writing tools, the attack surface for indirect injection is negligible. Evidence: Ingestion points: DESIGN.md, user brief; Boundary markers: absent; Capability inventory: none; Sanitization: absent.
Audit Metadata